Research ArticleMultimode Coordination Control of a Hybrid Active Suspension
Fa-Rong Kou Dong-Dong Wei and Lei Tian
School of Mechanical Engineering Xirsquoan University of Science and Technology Xirsquoan Shaanxi China
Correspondence should be addressed to Dong-Dong Wei 1462739700qqcom
Received 1 July 2018 Revised 23 October 2018 Accepted 11 November 2018 Published 16 December 2018
Academic Editor Francesco Braghin
Copyright copy 2018 Fa-Rong Kou et al 0is is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited
In order to effectively realize the damping control and regenerative energy recovery of vehicle suspension a new kind of hybridactive suspension structure with the ball screw actuator and magnetorheological (MR) damper is put forward Firstly for theanalysis of the suspension performance a quarter dynamic model of vehicle hybrid suspension is established and at the same timethe mathematical models of MR damper and ball screw actuator are founded Secondly the active mode with damping switchingcontrol of the hybrid suspension and the semiactive mode with feedback adjustment of the electromagnetic damping force of thehybrid suspension are analyzed 0en the multimode coordinated control system of the hybrid suspension is designed Under thecyclic driving condition the damping performance and energy consumption characteristics of the hybrid suspension aresimulated by MATLABSimulink software Finally the bench tests of the hybrid suspension system are done 0e simulation andexperimental results show that compared with passive suspension the root mean square of the sprung mass acceleration of thehybrid suspension with the active mode and semiactive mode is respectively reduced by 39 and 16 under the random road0e damping effect of the hybrid suspension system is obvious
1 Introduction
A controllable actuator is used to replace the correspondingcomponents of original passive suspension which is namedas an active suspension Active suspension can adjust sus-pension stiffness and damping according to changes in thecurrent road conditions to improve vehicle riding comfortand handling stability [1 2] However the active suspensionactuator needs to consume a large amount of external energyand it reduces the economic performance of the vehicle[3ndash5]
In recent years domestic and overseas scholars have beenstudying how to reduce the energy consumption of activesuspension from two aspects On the one hand the re-generative energy active suspension structures are used torecover vibration energy which is used for the active control ofthe suspension For example Nakano et al [6 7] used twolinear motor actuators for active control and vibration energyrecovery respectively and as a result the active suspensionsystem was self-powered Huang et al [8 9] proposed anew type of energy-regenerative electromagnetic suspensionstructure with a parallel-type ball screw actuator which in-cluded a fully active mode and semiactive mode Huang et al
[10] designed a regenerative energy active suspension systemand analyzed the power conversion process of the suspensionsystem and the simulation results show that under regener-ative energy and active mode switching control the suspen-sion system was self-powered However when the regenerativeenergy active suspension is used to recover vibration energy ofthe suspension the active control of the suspension cannot berealized so the riding comfort and handling and stability ofvehicle will be reduced
On the other hand domestic and overseas scholarsresearched how to reduce the energy consumption of activesuspensions by improving the suspension structures andoptimizing the structural parameters of active suspensionsFor example Ebrahimi et al [11] proposed a hybrid elec-tromagnetic shock absorber which was applied to an activesuspension developed a prototype of the shock absorberand carried out the experimental researches then the testsshow that the electromagnetic shock absorber was beneficialto reduce the energy consumption of the active suspensionTang et al [12] proposed an active suspension structure withparalleled three-gear variable damper and designed theswitching control strategy of the damper and the simulationresults show that for the suspension structure the energy
HindawiShock and VibrationVolume 2018 Article ID 6378023 16 pageshttpsdoiorg10115520186378023
consumption under the active control mode was reducedWang et al [13 14] proposed a hybrid suspension structurewith parallel damping variable shock absorber designed thevariable damping of the suspension in different workingmodes and carried out the simulations and experimentalresearch studies the results show that for the suspensionstructure the energy consumption under the active controlmode was reduced However for the research studies on theactive suspension structure and the structural parametersoptimization the energy consumption of the variabledamping shock absorber is not taken into account and theimpacts of the energy consumption on the economic per-formance of the suspension system are not analyzedMoreover under different vehicle speeds the impacts ofdifferent damping values of the variable damping shockabsorber on vehicle ride comfort handling and stability andeconomic performance are not analyzed
For the MR damper the continuous variable character-istics of the damping coefficient can be achieved by con-trolling the magnetic field strength of the MR fluid [15ndash17] Anew kind of hybrid active suspension structure with ball screwactuator and MR damper is put forward Moreover in orderto ensure high response speed of the hybrid active suspensiona sky-hook algorithm which is easy to operate fast and robustin response is applied [18ndash20] And the semiactive modeand active mode of the hybrid suspension are analyzed underthe class B road surface At the same time the multimodecoordinated control system of the hybrid suspension isdesigned 0en the damping performance and energy con-sumption characteristics of the hybrid suspension under thecyclic driving condition are simulated and the bench tests ofthe hybrid suspension system are done
2 Structure and Principle of the HybridActive Suspension
0e structure of the hybrid active suspension system isshown in Figure 1 It is mainly composed of spring MRdamper ball screw actuator controller battery corre-sponding signal detection device and so on
0e controller performs semiactive control or activecontrol of the hybrid suspension system by detecting andjudging the relevant signals When the hybrid suspension issemiactively controlled the controller detects the sprung massacceleration in real time by the sprung mass accelerationsensor and the semiactive control force of the hybrid sus-pension is obtained through the sky-hook algorithm 0econtroller adjusts the controllable current to change themagnetic field strength of the MR fluid 0en the outputdamping force of the MR damper is changed and the sem-iactive control of the suspension is realized At the same timethe ball screw actuator transforms the up and down motion ofthe suspension into the rotation of the motor and then themotor generates electric energy which is stored in the battery
When the hybrid suspension is actively controlled thecontroller detects the sprung mass acceleration in real time bythe sprung mass acceleration sensor and the active controlforce of the hybrid suspension is obtained through the sky-
hook algorithm 0e controller adjusts the controllable currentto change the motor output torque 0e motor transforms therotation motion into the up and down motion and the activecontrol of the suspension is realized At this time the energyconsumed by the ball screw actuator is supplied by the batteryAnd when the hybrid suspension is actively controlled thecontroller detects the running speed of vehicle and controls theinput current of the MR damper according to different vehiclespeed values It makes the MR damper produce a dampingvalue that matches the vehicle speed In the semiactive andactive control of the suspension the energy consumed by theMR damper is all provided by the battery
3 Modeling of Hybrid Active SuspensionDynamic Model
31 Dynamic Model of Hybrid Active Suspension In thispaper a quarter vehicle dynamic model of the hybrid activesuspension is established and shown in Figure 2
Based on Newtonrsquos laws of motion the dynamic motionequations for the quarter vehicle suspension can beexpressed as
ms eurox2 + ks x2 minusx1 1113857 + cs _x2 minus _x1 1113857 F
mu eurox1 minus ks x2 minusx1 1113857minus cs _x2 minus _x1 1113857 + kt x1 minus z 1113857 minusF1113896
(1)
0e state variable and the output vector are selected asfollows
X x2 minusx1 _x2 x1 minus z _x11113858 1113859T
Y eurox2x2 minusx1kt x1 minus z 1113857 _x11113858 1113859T
(2)
where ms is sprung mass mu is unsprung mass ks is springstiffness coefficient F is control force of suspension (especiallyFb is semiactive control force and Fz is active control force) ktis tire stiffness coefficient z is displacement of road input x2 isdisplacement of sprung mass x1 is displacement of unsprungmass and cs is damping coefficient of MR damper
0en the state-space equations of suspension system canbe described as follows
_X AX + BU
Y CX +DU
⎧⎨
⎩ (3)
Sprung mass acceleration sensor
SpringBall screw actuator
Unsprung mass acceleration sensor
Sprung massSuspension controller
Battery
Vehicle speed sensor
Road surface Wheel
MR damperUnsprung mass
Figure 1 Structure of the hybrid active suspension
2 Shock and Vibration
where AΒC and D are the state matrix input matrixoutput matrix and transfer matrix respectively When thecontrol input force F is 0 it becomes passive suspension
A filtered white noise is adopted as the road surface inputmodel as follows
_z(t) minus2πf0z(t) + 2πG0u
1113968ω(t) (5)
where G0 is the coefficient of road irregularity f0 is lowercutoff frequency u is vehicle speed and ω(t) is unit whitenoise
0e simulation parameters of the hybrid suspension arelisted in Table 1
32 MR Damper Mathematical Model Ignoring the frictionand fluid inertia of the MR fluid the damping force model ofthe MR damper under mixed operation mode is given asfollows [21]
Fa minus24ηA2
pl
bh3 +2ηblh
1113888 1113889 _x2 minus _x1 1113857minus4lAp
η+ 2bl1113888 1113889τy (6)
where Fa is MR damper output damping force η isfluid dynamic viscosity l is the working plate length his the working plate distance τy is MR fluid shear stressAp is piston effective area and b is the working platewidth
According to formula (6) the damping force of the MRdamper includes the viscous damping force Fn which has afunction relationship with the piston speed of the MRdamper and coulomb damping force Fk which has afunction relationship with the control current of the MRdamper So formula (6) can be transformed into
Fa minusce _x2 minus _x1 1113857 + a1I2k + a2I
2k + a31113872 1113873sgn _x2 minus _x1 1113857 (7)
where ce a1 a2 and a3 are polynomial coefficients and Ik iscontrol current of the MR damper
From formulae (6)sim(7) because the viscous dampingforce Fn does not consume energy it can be equivalent to thedamping force produced by the traditional hydraulic shockabsorber 0e variable damping force of the hybrid activesuspension is the coulomb damping force Fk which isachieved by adjusting Ik
In the semiactive control of the hybrid suspension theinstantaneous energy consumption power and the energyconsumption of the MR damper are expressed as
Pb Fk middot_x2 minus _x1 1113857
ηb (8)
Wb 1113946t
0Pb dt (9)
where ηb is work efficiency of MR damper Pb is the MRdamper instantaneous energy consumption power of thehybrid suspension in semiactive control and Wb is the MRdamper energy consumption of the hybrid suspension insemiactive control
In the active control of the hybrid suspension the outputdamping force of the MR damper is expressed as
ms
mu
cs
F
z
ks
kt
x1
x2
Figure 2 Dynamic model of 2 DOF hybrid active suspension
Shock and Vibration 3
cs c0 + ck
c0 Fn
_x2 minus _x1( )
ck Fk
_x2 minus _x1( )
(10)
where c0 is viscous damping of the MR damper and ck isvariable damping of the MR damper
In the active control the instantaneous energy con-sumption power and the energy consumption of the MRdamper are expressed as
Pc cs minus c0( ) middot _x2 minus _x1( )2
ηb (11)
Wc intt
0Pc dt (12)
where Pc is the MR damper instantaneous energy con-sumption power of the hybrid suspension in active controlandWc is theMR damper energy consumption of the hybridsuspension in active control
e single rod MR damper is used in this paper and it isshown in Figure 3
By carrying out the characteristics test of the MR damperand analyzing the test data the relation diagrams of thedamping force-velocity curves of the MR damper are ob-tained and shown in Figure 4
33 Ball Screw Actuator Mathematical Model
331 e Characteristics of Ball Screw Actuator e ballscrew actuator is used not only to realize the active control ofthe hybrid suspension but also to recover the regenerativeenergy of the hybrid suspension the characteristics of whichhave a great inshyuence on the performance of the hybridsuspension
e characteristics of the ball screw actuator is mainlyaected by the back-EMF coecient ke and electric torquecoecient kT of the motor but the nonlinear characteristicsof the motor make ke and kT vary with the speed of themotor [22 23] It is necessary to gain the relationship be-tween ke kT and the motor speed by the motor test if aprecise mathematical model of the ball screw actuator isestablished
e prototype of the ball screw actuator is shown inFigure 5 It is made up of brushless DC motor ball screwupper and lower ears force sensor and so on
In this paper the peak value of counter electromotiveforce of the motor at dierent rotating speeds is measured bythe test At dierent rotating speeds the tting analyses of thepeak value are carried out and the result is shown in Figure 6
e peak of counter electromotive force of themotor andthe back-EMF coecient meet the following relationships
ke Vmax
22
radicn (13)
where Vmax is the peak of counter electromotive force and nis motor rotating speed
Table 1 Simulation parameters of the hybrid suspension
Parameters ValuesLower cuto frequency (Hz) 01Road irregularities coecient (m3) 64 times 10minus6
Sprung mass (kg) 38Unsprung mass (kg) 210Spring stiness coecient (kNm) 22Tire stiness coecient (kNm) 200
Upper ear
MR damper Connecting rod Lower ear
Force sensor
Figure 3 e MR damper prototype
ndash01 ndash005 0 005 01ndash600
ndash400
ndash200
0
200
400
600
Vibration speed (mmiddotsndash1)
Dam
ping
forc
e (N
)
02 A04 A06 A
08 A10 A
Figure 4 e damping force-velocity curves of the MR damper
Upper ear Force sensor
Ball screwBrushless DC motor Lower ear
Figure 5 e ball screw actuator
4 Shock and Vibration
e vibration velocity of the hybrid suspension and themotor rotating speed meet the following relationships
n ΔvLmiddot 60 (14)
Δv _x2 minus _x1 (15)
where Δv is the suspension vibration velocity and L is ballscrew lead
At dierent suspension vibration velocity the motorback-EMF coecient can be obtained from formulae(13)sim(15) and the tting curve of the motor back-EMFcoecient is shown in Figure 7
In Figure 7 the tting relationship between ke and Δv isexpressed as
ke 00299Δv3 minus 00542Δv2 + 00253Δv + 00049 (16)
en kT and ke meet the following relationships
kT 30πke (17)
e relationship between kT and Δv can be obtainedfrom formulae (16) and (17)
kT 02855Δv3 minus 05176Δv2 + 02416Δv + 00468 (18)
When testing the active output force of the ball screwactuator the ball screw actuator is powered by the same 72Vconstant voltage source as the battery pack terminal voltageBy adjusting PWM duty ratio the input voltage and currentof the motor are controlled by the controller and as a resultthe output force of the motor is gained e relationshipbetween the active output force of the ball screw actuatorand duty ratio is shown in Figure 8
From Figure 8 it can be seen that the ball screw actuatorhas good active output force characteristics and the non-linear relationship between the output force and the dutyratio is consistent with the nonlinear characteristics of themotor
332 Ball Screw Actuator Mathematical Model When themotor inductance is ignored in the active control of thehybrid suspension input voltage E and the output torqueTmof the motor meet the following relationships
E u + Izr (19)
Tm kT middot Iz (20)
u 60 _x2 minus _x1( )
Lmiddot ke (21)
Tm F middot L2π
(22)
ke kT middot π30
(23)
400 800 1200 1600 2000 2400 2800 3200 3600 40000
10
20
30
40
50
60
70
Rotating speed (rmin)
The p
eak
valu
e of c
ount
erel
ectr
omot
ive f
orce
(V)
The peak of counter electromotive forceTest fitting curve
Figure 6 e peak value curve of counter electromotive force ofthe motor
008 020 032 044 056 068 08000055
00060
00065
00070
00075
00080
00085
Suspension vibration speed (ms)
Back
-EM
F co
effic
ient
(Vr
middotmin
ndash1)
Test value data pointsTest fitting curve
Figure 7 e motor back-EMF coecient curve
ndash100 ndash50 0 50 100ndash600
ndash400
ndash200
0
200
400
600
Duty cycle ()
Act
uato
r act
ive o
utpu
t for
ce (N
)
Actuator active output forceTest fitting curve
Figure 8 Testing curve of the actuator active output force
Shock and Vibration 5
where u is induced electromotive force Iz is motor currentand r is internal resistance of the motor
In the active control of the hybrid suspension the in-stantaneous energy consumption power of the ball screwactuator can be expressed as
Pz E middot Iz
ηz (24)
0e instantaneous energy consumption power andconsumption energy can be obtained from formulae(19)sim(24) as follows
Pz _x2 minus _x1 1113857 middot F + FL2πkT 1113857
2middot r1113960 1113961
ηz (25)
Wz 1113946t
0Pz dt (26)
where Pz is motor instantaneous energy consumptionpower ηz is transfer efficiency of ball screw actuator andWzis motor energy consumption
In the energy regeneration of the hybrid suspension theelectromagnetic damping force generated by the ball screwactuator is expressed as
Fs minus2πkT
L1113888 1113889
2
middot_x2 minus _x1 1113857
R + rηz (27)
where Fs is the electromagnetic damping force and R isexternal resistance of the motor
0e instantaneous energy-regenerative power and regen-erative energy of the ball screw actuator can be expressed as
Pk 2πke
L1113888 1113889
2
middot_x2 minus _x1 1113857
2
R + rηz (28)
Wk 1113946t
0Pk dt (29)
where Pk is instantaneous energy-regenerative power andWk is regenerative energy
4 Multimode Coordination Control of HybridActive Suspension
41 -e Active Mode of Damping Switching Control 0esprung mass acceleration is the main evaluation index ofvehicle riding comfort and the dynamic tire load is closelyrelated to vehicle handling and stability An active controlmodel of the hybrid suspension is established by MATLABSimulink software to simulate and analyze the influence ofthe variable damping for the hybrid active suspension onvehicle riding comfort and handling and stability at differentvehicle speeds During the simulation the range of vehiclespeed v is 0ndash120 kmh and the vehicle speed is taken every10 kmh 0e variable damping range of the suspension is200ndash2000 Nmiddotsm and the value of the variable damping istaken every interval 100 Nmiddotsm 0e simulation time is 10 sand the value of r is 05 Ω 0e value of ηz is 097 and thevalue of ηb is 098 0e value of csky is 2000 Nmiddotsm
When the vehicle speed is 30 kmh and 100 kmh re-spectively the RMS of the sprung mass acceleration of thevehicle (aw) and the RMS of the dynamic tire load (DTLrms)change with the variable damping of the hybrid activesuspension as shown in Figures 9 and 10
From Figures 9 and 10 it can be seen that when thevehicle speed is 30 kmh the variable damping values thatmake aw of vehicle and DTLrms minimum are 400 Nmiddotsm and1000 Nmiddotsm respectively and when the vehicle speed is100 kmh the variable damping values that make aw ofvehicle and DTLrms minimum are 500 Nmiddotsm and 1100 Nmiddotsm respectively 0erefore at a certain vehicle speed thevariable damping of the hybrid active suspension cannotmake the best of the vehicle riding comfort and handling andstability at the same time
Suspension performance indexes include the sprung massacceleration suspension working space and dynamic tireload In this paper in order to balance vehicle riding comfortand handling and stability when choosing the variabledamping values of the hybrid suspension in the active controlfor the sprung mass acceleration suspension working spaceand dynamic tire load of the hybrid active suspension thequantitative normalizations and comparative analyses aredone 0at is at the same vehicle speed compared with thepassive suspension the improvement amplitudes of eachperformance index of the hybrid active suspension aremultiplied by different quantification factors and summedAnd the larger the sum the better the dynamic performanceof vehicle Among them the quantification factors of aw theRMS value of suspension working space (SWSrms) andDTLrms are 1 02703 and 01443 respectively [24 25] Andwhen the vehicle speed is 30 kmh and 100 kmh respectivelythe dynamic performance and the active control energyconsumption of the hybrid active suspension change with thevariable damping as shown in Figures 11 and 12
From Figures 11 and 12 when the vehicle speed is 30 kmh the variable damping value of the hybrid active suspensionis 800 Nmiddotsm which makes the vehicle dynamic performancethe best and the active energy consumption the least Whenthe vehicle speed is 100 kmh the variable damping value ofthe hybrid active suspension is 1000 Nmiddotsm which makes thevehicle dynamic performance the best and active energyconsumption the least 0erefore when the vehicle speed is30 kmh and 100 kmh respectively the optimal dampingvalues of the hybrid active suspension are 800 Nmiddotsm and1000 Nmiddotsm respectively When the vehicle speed is 0ndash120 kmh the optimal damping values of the hybrid activesuspension at different vehicle speeds are shown in Figure 13If the damping value which makes the vehicle dynamicperformance the best is different from the damping valuewhich makes the active control energy consumption the leastthe damping value which makes the vehicle dynamic per-formance the best is selected as the optimal damping value ofthe hybrid active suspension at the vehicle speed
When vehicle is in an accelerating or decelerating stateits speed changes rapidly and the range of change is wide sothe vehicle speed value is not easily detected in real time andin order to reduce the energy consumption of the hybridsuspension active controlled and to improve the vehicle
6 Shock and Vibration
riding comfort and handling and stability a variabledamping switching control strategy of the hybrid activesuspension is designed as follows
Fz minus csky middot _x2cs c0 + ck
_v 0
Fz minus csky middot _x2cs c0
_vne 0
(30)
where _v is vehicle acceleration and csky is sky-hook coecient
42 e Semiactive Mode of Feedback Adjustment of Elec-tromagnetic Damping Force MR damper can eectivelyperform semiactive control at ( _x2 minus _x1) _x2 gt 0 so the idealsemiactive control state for hybrid suspension is
Fb minus csky middot _x2 _x2 minus _x1( ) _x2 gt 0
0 _x2 minus _x1( ) _x2 le 0
(31)
From equation (31) the ideal semiactive control force ofthe hybrid suspension minus csky middot _x2 is only related to _x2 when
Vehicle dynamics performanceEnergy consumption of hybrid suspension
Figure 12 Relationship between the variable damping and vehicledynamic performance at 100 kmh speed
20 40 60 80 100 120750
800
850
900
950
1000
1050
Vehicle speed (kmh)
Shoc
k ab
sorb
er d
ampi
ngco
effic
ient
(Nmiddots
m)
Figure 13 Suspension optimal damping at dierent speeds
Shock and Vibration 7
the sky-hook coecient csky is constant However at thistime the ball screw actuator as a power feeding devicegenerates the electromagnetic damping force Fs and acts onthe suspension so that the actual semiactive control force ofthe suspension is dierent from the ideal semiactive controlforce minus csky middot _x2 In this paper the semiactive control modelof the hybrid suspension is established and the changeeects of the dierent output forces on the vehicle ridingcomfort and handling stability are analyzed by MATLABSimulink software e simulation speed is 70 kmh thesimulation time is 5 s and the value of R is 075 Ω edamping comparison of the hybrid suspension in semiactivecontrol is shown in Figure 14
Figure 14 shows that compared with the ideal semi-active control force the actual semiactive control force ofthe hybrid suspension shyuctuates violently and the abso-lute value of the actual semiactive control force is greaterthan the absolute value of the ideal semiactive control force|csky middot _x2| at certain times And a drastic change in theactual semiactive control force makes the suspension notreach ideal semiactive control eect Using electromag-netic damping force feedback adjustment to reduce thedierence between the ideal semiactive control forceminus csky middot _x2 and the actual semiactive control force themethod is as follows
When |csky middot _x2|gt |Fs| the semiactive control force of thehybrid suspension is provided by both the MR damper andthe ball screw actuator and at this point the controllerinputs a controllable current Ik to the MR damper so thatthe Fk output by the MR damper is minus csky middot _x2 minusFs Andwhen |csky middot _x2|le |Fs| the semiactive control force of thehybrid suspension is the Fs which is output by the ball screwactuator and at this point there is no controllable current Ikinput to theMR damper and the function of theMR damperis equivalent to a traditional shock absorber erefore thesemiactive control of the hybrid suspension does not havethe dead zone of traditional electromagnetic semiactivesuspension which helps to improve the semiactive controleect of the hybrid suspension
When there is feedback adjustment the semiactivecontrol force of the hybrid suspension is
Fb Fs Fs
∣∣∣∣∣∣∣∣ge csky middot _x2∣∣∣∣∣
∣∣∣∣∣
minus csky middot _x2 Fs∣∣∣∣∣∣∣∣lt csky middot _x2∣∣∣∣∣
∣∣∣∣∣
(32)
When there is feedback adjustment the Fk output by theMR damper is
Fk 0 Fs
∣∣∣∣∣∣∣∣ge csky middot _x2∣∣∣∣∣
∣∣∣∣∣
minus csky middot _x2 minusFs Fs∣∣∣∣∣∣∣∣lt csky middot _x2∣∣∣∣∣
∣∣∣∣∣
(33)
From equations (31)sim(33) when the electromagneticdamping force feedback adjustment is used the |Fk| outputby theMR damper decreases and when |Fk| decreases it canbe known from equations (8) and (9) that the energyconsumption of the MR damper decreases with it
e comparison of the semiactive control force of thehybrid suspension with or without the electromagneticdamping force feedback adjustment is shown in Figure 15
From Figure 15 the RMS of the ideal semiactive controlforce of the hybrid suspension is 3276 N and when there isno electromagnetic damping force feedback adjustmentthe RMS of the actual semiactive control force of thesuspension is 4041 N and the dierence between the actualsemiactive force of the suspension and the ideal semiactiveforce is 2335 When there is electromagnetic dampingforce feedback adjustment the RMS of the actual semi-active control force of the suspension is 3593 N and thedierence between the actual semiactive force of the sus-pension and the ideal semiactive force is 968 ereforewhen there is electromagnetic damping force feedbackadjustment the actual semiactive control force of thesuspension has a smaller shyuctuation amplitude whichhelps to improve the semiactive control eect of the hybridsuspension
e dynamic responses of the hybrid suspension with orwithout electromagnetic damping force feedback adjust-ment are shown in Figure 16 Among them the damper ofthe passive suspension is the original damper of the vehicleand its damping value is 1600Nmiddotsm
Table 2 shows the response RMS values of the hybridsuspension in semiactive control
From Table 2 compared with the passive suspensionwhen there is electromagnetic damping force feedbackadjustment aw SWSrms and DTLrms of the hybrid sus-pension are reduced by 1698 432 and 1068 re-spectively and compared with the nonfeedback semiactivecontrol when the feedback semiactive control is performedaw SWSrms and DTLrms of the hybrid suspension are re-duced by 252 863 and 671 respectively
From equations (9) and (29) the total system energy ofthe hybrid suspension in semiactive control is
W1 Wk minusWb (34)
whereW1 is the total system energy of the hybrid suspensionin semiactive control
0 1 2 3 4 5ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Hyb
rid su
spen
sion
forc
e (N
)
072 076 080
300
600
Ideal semiactive forceActual semiactive forceElectromagnetic damping force
Figure 14 Damping comparison of the hybrid suspension insemiactive control
8 Shock and Vibration
0 1 2 3 4 5ndash10
ndash5
0
5
10
15
Time (s)
Spru
ng m
ass a
ccel
erat
ion
(ms
2 )
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(a)
0 1 2 3 4 5ndash004
ndash002
000
002
004
006
008
Time (s)
Susp
ensio
n w
orki
ng sp
ace (
m)
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(b)
0 1 2 3 4 5ndash3000
ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Dyn
amic
tire
load
(N)
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(c)
Figure 16 e dynamic responses of the hybrid suspension in semiactive control (a) e response curves of sprung mass acceleration (b)e response curves of suspension working space (c) e response curves of dynamic tire load
0 1 2 3 4 5ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Hyb
rid su
spen
sion
forc
e (N
)
072 076 080
300
600
Ideal semiactive forceWithout feedback semiactive forceWith feedback semiactive force
Figure 15 Semiactive force of the hybrid suspension
Shock and Vibration 9
From equation (33) when the hybrid suspension sem-iactive controlled the curves of the total system energychange over time are shown in Figure 17
From Figure 17 when there is the nonfeedback semiactivecontrol the total system energy of the hybrid suspension is60 J And when there is the feedback semiactive control thetotal system energy of the hybrid suspension is 307 J
43 -e Design of Multimode Coordination ControllerWhen the hybrid suspension is actively controlled thesuspension has good vibration isolation performance buthigh energy consumption And when the hybrid suspensionis semiactively controlled the suspension has good eco-nomic performance but the control has limitations Con-sidering that the ball screw actuator can realize active controlof the hybrid suspension in any suspension state a multi-mode coordinated control strategy of the hybrid suspensionis designed
minus csky middot _x2 Fs1113868111386811138681113868
1113868111386811138681113868gt csky middot _x2
11138681113868111386811138681113868
11138681113868111386811138681113868
⎧⎪⎨
⎪⎩
cs c0
⎧⎪⎪⎪⎨
⎪⎪⎪⎩
_x2 minus _x1 1113857 _x2 gt 0
F Fz minus csky middot _x2
cs c0 + ck1113896 _x2 minus _x1 1113857 _x2 le 0
⎧⎪⎪⎪⎪⎪⎪⎪⎪⎪⎨
⎪⎪⎪⎪⎪⎪⎪⎪⎪⎩
(35)
0e frame diagram of the multimode coordinated controlstrategy of the hybrid suspension is shown in Figure 18
From equations (9) (12) (26) and (29) the total systemenergy of the hybrid suspension in multimode coordinatedcontrol is
W2 Wk minusWb minusWZ minusWc (36)
whereW2 is the total system energy of the hybrid suspensionin multimode coordinated control
A hybrid active suspension simulation model is estab-lished by using MATLABSimulink software From Fig-ure 11 when the hybrid suspension is actively controlled theenergy consumption of the MR damper is mainly affected bythe vehicle speed In order to verify the vibration isolationperformance and energy consumption performance of thehybrid active suspension in the cyclic driving conditionsthis article simulates vehicle urban and suburban conditionswhich is based on GBT 19233-2003 ldquoLight Vehicle FuelConsumption Test Methodrdquo among them the urban con-ditions include four cycle units each cycle time is 195 s thesuburban conditions include one cycle unit and the cycletime is 400 s [26] And the schematic diagram of the urbanconditions unit is shown in Figure 19
In order to simulate the vehicle acceleration signal avehicle speed variation model in different cycle units isestablished by using the signal builder function module inSimulink software On this basis the vehicle accelerationmodel in different cycle units is obtained and the vehicleacceleration model can be used as the switching controlmodel for the variable damping when the hybrid suspensionis actively controlled 0e shock absorbers of active sus-pension and passive suspension adopt the original damper ofvehicle and the damping value c1 is 1600 Nmiddotsm Howeverbecause of the long simulation time the dynamic responsecurves of the hybrid active suspension in urban and sub-urban conditions cannot be displayed Figure 20 shows thedynamic response curve of the hybrid active suspension inthe 145ndash175 s in the urban circulation unit among themvehicle in the 145ndash155 s is in a constant speed vehicle in the155ndash163 s is in the deceleration state and vehicle in the163ndash175 s is in other constant speed
0e dynamic response RMS values of the hybrid sus-pension in urban and suburban circulation units are shownin Tables 3 and 4 respectively
From Figure 20 and Tables 3 and 4 the dynamic re-sponses are good when the hybrid suspension is multimodecoordinated control When in the urban circulation unitcompared with the passive suspension aw SWSrms andDTLrms of the hybrid suspension are reduced by 3943225 and 2081 respectively Compared with the activesuspension aw and DTLrms of the hybrid suspension arereduced by 1322 and 840 respectively but SWSrms isincreased by 805 When in the suburban circulation unitcompared with the passive suspension aw SWSrms andDTLrms of the hybrid suspension are reduced by 3916314 and 1955 respectively Compared with the activesuspension aw and DTLrms of the hybrid suspension arereduced by 1199 and 732 respectively but SWSrms isincreased by 845
0e system energy of the hybrid suspension in urban andsuburban circulation units are shown in Figures 21 and 22respectively
Table 2 0e response root mean square values of the hybridsuspension in semiactive control
Without feedback semiactive controlWith feedback semiactive control
Figure 17 0e system energy of the hybrid suspension in semi-active control
10 Shock and Vibration
en the system energy values of the hybrid suspensionin urban and suburban circulation units are shown inTable 5
From Table 5 when in the urban circulation unit thesystem energy of the active suspension and hybrid sus-pension is minus2165 J and 38 J respectively And when in thesuburban circulation unit the system energy of the activesuspension and hybrid suspension is minus15071 J and minus122 Jrespectively e entire operation cycle includes 4 urbancycle units and 1 suburban cycle unit so the pure energy ofthe active suspension system during the entire operationcycle is minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid suspensionsystem basically realizes energy self-powered in theory
5 Test and Analysis
In order to verify the damping eect of the hybrid sus-pension system in active mode and semiactive mode ahybrid suspension vibration test system is designed as
Suspension controllerSpeed of sprung mass x2
Speed of unsprung mass x1Vehicle acceleration v
Vehicle suspension
Ball screw actuator
Output Fk
Active control force Fz
Ball screw actuator
Output Iz
Output Ik
Electromagnetic damping force Fs
Ideal semiactive control force FL
Output semiactive controlforce Fb = Fs
Output semiactive controlforce Fb = FL = Fk + Fs
x2 ndash x1
(x2 ndash x1) x2 gt 0
(x2 ndash x1) x2 le 0
Output Ik
MR damper
Adjust the duty cycle in real timeaccording to the relationshipbetween active output force
and duty cycle
No controllablecurrent output
MR damper
v ne 0
According to vand ∆v
v = 0
MR damper
Adjustable damping cs = c0 + ck
ndashcsky middot x2
ndashcsky middot x2
|Fs| ge |FL|
|Fs| lt |FL|
Figure 18 e frame diagram of the multimode coordinated control strategy of the hybrid suspension
Vehicle shifting
00
15
10
30
30
20
45
40
50
60
60
75 90 105 120Time (s)
Veh
icle
spee
d (k
mh
)
135 150 165 180 195
Figure 19 e schematic diagram of the urban conditions unit
Shock and Vibration 11
shown in Figure 23 During the test the MR damper has nocontrollable current input in the active control mode of thehybrid suspension and its damping value is always 800Nmiddotsm
e passive suspension semiactive suspension and activesuspension all adopt the original damper of the vehicle andthe damping value c1 is 1600Nmiddotsm
Figure 20e dynamic responses of the hybrid suspension (a)e response curves of sprungmass acceleration (b)e response curves ofsuspension working space (c) e response curves of dynamic tire load
12 Shock and Vibration
Because of the limitation of test conditions only thesprung mass acceleration dynamic response of the hybridsuspension is measured in this test e dynamic response ofthe sprung mass acceleration of the hybrid suspension inactive mode under random road is shown in Figure 24 epower spectrum of the sprung mass acceleration of thehybrid suspension in active mode is shown in Figure 25
Table 6 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in active mode
From Table 6 compared with passive suspension aw ofthe hybrid suspension in active mode is reduced by 3945and compared with active suspension aw of the hybridsuspension in active mode is reduced by 1432 And fromFigure 25 when the hybrid suspension is actively controlledcompared with passive suspension the vibration isolationperformance of the suspension in the low frequency andlow-frequency resonance areas is similar to that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withactive suspension the vibration isolation performance of thesuspension in the low frequency high frequency and res-onance regions is slightly better
e dynamic response of the sprung mass acceleration ofthe hybrid suspension in semiactive mode under randomroad is shown in Figure 26 e power spectrum of thesprung mass acceleration of the hybrid suspension insemiactive mode is shown in Figure 27
Table 7 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in semiactive mode
From Table 7 compared with passive suspension aw ofthe hybrid suspension in semiactive mode is reduced by1642 Compared with semiactive suspension aw of thehybrid suspension in active mode is reduced by 307 Andfrom Figure 27 when the hybrid suspension is semiactivelycontrolled compared with passive suspension the vibration
Table 3 e dynamic response root mean square values of thehybrid suspension in urban circulation unit
Figure 21 e system energy of the hybrid suspension in urbancirculation units
0 100 200 300 400ndash20000
ndash15000
ndash10000
ndash5000
0
5000
Time (s)
Susp
ensio
n sy
stem
ener
gy (J
)
Active suspensionHybrid suspension
Figure 22e system energy of the hybrid suspension in suburbancirculation units
Table 5 e system energy values of the hybrid suspension (J)
Cycle unit Active suspension Hybrid suspensionUrban minus2165 38Suburban minus15071 minus122
Figure 23 e test system of the hybrid active suspension
Shock and Vibration 13
isolation performance of the suspension in the low frequencyand low-frequency resonance areas is worse than that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withsemiactive suspension the vibration isolation performance ofthe suspension in the low frequency high frequency andresonance regions is slightly better
e energy consumption power and energy-regenerativepower of the hybrid suspension are shown in Figure 28
From Figure 28 in the active mode the average powerconsumption of the ball screw actuator is 2561W In thesemiactive mode the average regenerative power of the ballscrew actuator is 2696W and the average of theMR damperenergy consumption power is 178W erefore the pureaverage power consumption of the hybrid suspension isminus043W and the test results are basically consistent with thesimulation
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spr
ung
mas
s acc
eler
atio
n (m
s2 )
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 24 e dynamic response of the sprung mass accelerationof the hybrid suspension in active mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 25 e power spectrum of the sprung mass acceleration ofthe hybrid suspension in active mode
Table 6 e sprung mass acceleration RMS of the hybrid sus-pension in active mode
Indicators Passivesuspension
Activesuspension
Hybridsuspension
aw (ms2) 13378 09423 08074
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spru
ng m
ass a
ccel
erat
ion
(ms
2 )
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 26 e dynamic responses of the sprung mass accelerationof the hybrid suspension in semiactive mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 27e power spectrums of the sprungmass acceleration ofthe hybrid suspension in semiactive mode
Table 7 e sprung mass acceleration RMS of the hybrid sus-pension in semiactive mode
Indicators Passivesuspension
Semiactivesuspension
Hybridsuspension
aw (ms2) 13378 11536 11182
14 Shock and Vibration
6 Conclusion
(1) A ball screw actuator andMR damper are introducedinto vehicle suspension system and a new kind ofhybrid active suspension structure is put forwarde ball screw actuator is prototyped and thefunctional relationship between the back-EMF co-ecient the electromagnetic torque coecient of themotor and the suspension vibration speed is ob-tained by test analyses And the active output me-chanical properties of the ball screw actuator aretested and the results show that the actuator hasgood active output force characteristics
(2) e inshyuences of the variable damping value of thesuspension system on the riding comfort handlingand stability and energy consumption characteristicsof the hybrid suspension in the active control modeare analyzeden the optimal damping values of thehybrid suspension at dierent vehicle speeds aredesigned e eects of electromagnetic dampingforce on the actual semiactive force and the systemenergy of the suspension in the semiactive controlmode are analyzed and then the hybrid suspensionwith semiactive mode which has electromagneticdamping force feedback adjustment is designed Onthis basis a multimode coordinated control strategyfor the hybrid suspension is designed
(3) e damping performance and energy consumptioncharacteristics of the hybrid suspension under cyclicdriving condition are simulated by MATLABSimulink software and the results show that whenin the urban circulation unit compared with activesuspension aw and DTLrms of the hybrid suspensionare reduced by 1322 and 840 respectively Andwhen in the suburban circulation unit comparedwith the active suspension aw and DTLrms of thehybrid suspension are reduced by 1199 and 732respectively e pure energy of the active suspen-sion system during the entire operation cycle is
minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid sus-pension system basically realizes energy self-poweredin theory
(4) e eectiveness verication test of the hybrid sus-pension in active mode and semiactive mode controlis carried out and the results show that when thehybrid suspension is actively controlled comparedwith active suspension aw of the hybrid suspension inactive mode is reduced by 1432 When the hybridsuspension is semiactively controlled compared withsemiactive suspension aw of the hybrid suspension inactive mode is reduced by 307 e test and sim-ulation results are basically consistent and the testveries the correctness of the simulation
Data Availability
e data used to support the ndings of this study are in-cluded within the supplementary information les And thedata used to support the ndings of this study are availablefrom the corresponding author upon request
Conflicts of Interest
e authors declare that there are no conshyicts of interestregarding the publication of this paper
Acknowledgments
is work was supported by the National Natural ScienceFoundation of China (Grant no 51775426) Service LocalSpecial Program Support Project of Shaanxi ProvincialEducation Department (Grant no 17JF017) and XirsquoanScience and Technology Program Funding Project (Grantno 2017079CGRC042-XAKD007)
Supplementary Materials
e supplementary materials are this articlersquos experimentaldata including (1) MR damperrsquos characteristic test data (2)the motor counter electromotive forcersquos peak test data (3)the ball screw actuatorrsquos active output force test data (4) thehybrid suspensionrsquos sprung mass acceleration test data inactive mode (5) the hybrid suspensionrsquos sprung mass ac-celeration test data in semiactive mode and (6) the hybridsuspensionrsquos energy consumption power and energy re-generation power test data (Supplementary Materials)
References
[1] M Montazeri-Gh and O Kavianipour ldquoInvestigation of theactive electromagnetic suspension system considering hybridcontrol strategyrdquo Proceedings of the Institution of MechanicalEngineers Part C Journal of Mechanical Engineering Sciencevol 228 no 10 pp 1658ndash1669 2013
[2] H E Tseng and D Hrovat ldquoState of the art survey active andsemi-active suspension controlrdquo Vehicle System Dynamicsvol 53 no 7 pp 1034ndash1062 2015
0 1 2 3 4 50
20
40
60
80
Time (s)
Pow
er (W
)
Energy consumption power in active modeEnergy regenerative power in semiactive modeEnergy consumption power in semiactive mode
Figure 28 e energy consumption power and energy-re-generative power of the hybrid suspension
Shock and Vibration 15
[3] F Kou J Du Z Wang D Li and J Xu ldquoNonlinear modelingand coordinate optimization of a semi-active energy re-generative suspension with an electro-hydraulic actuatorrdquoChina Mechanical Engineering vol 28 no 14 pp 1701ndash17072017
[4] S H Zareh A Sarrafan A A A Khayyat and A ZabihollahldquoIntelligent semi-active vibration control of eleven degrees offreedom suspension system using magnetorheologicaldampersrdquo Journal of Mechanical Science and Technologyvol 26 no 3 pp 323ndash334 2012
[5] S A Chen X Li L J Zhao Y X Wang and Y B KimldquoDevelopment of a control method for an electromagneticsemi-active suspension reclaiming energy with varying chargevoltage in stepsrdquo International Journal of Automotive Tech-nology vol 16 no 5 pp 765ndash773 2015
[6] K Nakano Y Suda and S Nakadai ldquoSelf-powered activevibration control using a single electric actuatorrdquo Journal ofSound and Vibration vol 260 no 2 pp 213ndash235 2003
[7] K Nakano ldquoCombined type self-powered active vibrationcontrol of truck cabinsrdquo Vehicle System Dynamics vol 41no 6 pp 449ndash473 2004
[8] K Huang Y C Zhang F Yu and Y H Gu ldquoCoordinateoptimization for synthetical performance of electrical energy-regenerative active suspensionrdquo Journal of Shanghai JiaotongUniversity vol 43 no 2 pp 226ndash230 2009
[9] K Huang F Yu and Y C Zhang ldquoActive control of energy-regenerative electromagnetic suspension based on energy flowanalysisrdquo Journal of Shanghai Jiaotong University vol 45no 67 pp 1068ndash1073 2011
[10] D S Huang J Q Zhang Y L Liu L Yi and X Y WangldquoPerformance of a novel energy-regenerative active suspen-sion systemrdquo Journal of Chongqing University (English Edi-tion) vol 14 no 3 pp 109ndash118 2015
[11] B Ebrahimi H Bolandhemmat M B Khamesee andF Golnaraghi ldquoA hybrid electromagnetic shock absorber foractive vehicle suspension systemsrdquo Vehicle System Dynamicsvol 49 no 1-2 pp 311ndash332 2011
[12] S Tang L Chen R Wang X Sun and D Shi ldquoResearch onoptimal control of active suspension based on damping multi-modal switchingrdquo Journal of Guangxi University (NaturalScience) vol 39 no 2 pp 300ndash307 2014
[13] R Wang X Ma R Ding X Meng and L Chen ldquoResearchof multi-mode switching control system for hybrid sus-pension based on model referencerdquo Transactions of theChinese Society for Agricultural Machinery vol 48 no 7pp 353ndash360 2017
[14] R Wang Y Qian R Ding X Meng and J Xie ldquoDesign andtests for damping-stiffness of a hybrid electromagnetic sus-pension based on LQGrdquo Journal of Vibration and Shockvol 37 no 3 pp 61ndash65 2017
[15] B Vanavil K K Chaitanya and A S Rao ldquoImproved PIDcontroller design for unstable time delay processes based ondirect synthesis method and maximum sensitivityrdquo Taylorand Francis vol 46 no 8 pp 1349ndash1366 2015
[16] I Mihai and F Andronic ldquoBehavior of a semi-active sus-pension system versus a passive suspension system on anuneven road surfacerdquo Mechanics vol 20 no 1 pp 64ndash692014
[17] D Ngoduy ldquoLinear stability of a generalized multi-anticipative car following model with time delaysrdquo Com-munications in Nonlinear Science and Numerical Simulationvol 22 no 1 pp 420ndash426 2015
[18] H Zhang E Wang F Min R Subash and C Su ldquoSkyhook-based semi-active control of full-vehicle suspension with
magneto-rheological dampersrdquo Chinese Journal of Mechan-ical Engineering vol 26 no 3 pp 498ndash505 2013
[19] F Kou ldquoAn experimental study on the dynamic character-istics of vehicle semi-active seat suspension with magneto-rheological damperrdquo Automotive Engineering vol 37 no 11pp 1346ndash1352 2015
[20] Z Feng S Chen and Y Liang ldquoAn experimental study on thedynamic characteristics of a megneto-rheological semi-activesuspensionrdquoAutomotive Engineering vol 35 no 1 pp 72ndash772013
[21] Z Li and L Xu A New Type of Magnetorheological Damperand Semi-Active coNtrol Design-eory Science Press BeijingChina 2012
[22] B L J Gysen J J H Paulides J L G Janssen andE A Lomonova ldquoActive electromagnetic suspension systemfor improved vehicle dynamicsrdquo IEEE Transactions on Ve-hicular Technology vol 59 no 3 pp 1156ndash1163 2010
[23] Q N Wang S S Liu W H Wang and H Wei ldquoStructuredesign and parameter matching of ball-screw regenerativedamperrdquo Journal of Jilin University (Engineering and Tech-nology Edition) vol 42 no 5 pp 1100ndash1106 2012
[24] S Chen R He and S Lu ldquoEvaluating system of reclaimingenergy suspension comprehensive performancerdquo Trans-actions of the Chinese Society for Agricultural Machineryvol 37 no 7 pp 14ndash18 2006
[25] Y Fan Control Research on Vehicle Suspension with Electro-Hydrostatic Actuator Xirsquoan University of Science and Tech-nology Xirsquoan China 2017
[26] Z Yu Automobile -eory China Machine Press BeijingChina 2009
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Navigation and Observation
International Journal of
Hindawi
wwwhindawicom Volume 2018
Advances in
Multimedia
Submit your manuscripts atwwwhindawicom
consumption under the active control mode was reducedWang et al [13 14] proposed a hybrid suspension structurewith parallel damping variable shock absorber designed thevariable damping of the suspension in different workingmodes and carried out the simulations and experimentalresearch studies the results show that for the suspensionstructure the energy consumption under the active controlmode was reduced However for the research studies on theactive suspension structure and the structural parametersoptimization the energy consumption of the variabledamping shock absorber is not taken into account and theimpacts of the energy consumption on the economic per-formance of the suspension system are not analyzedMoreover under different vehicle speeds the impacts ofdifferent damping values of the variable damping shockabsorber on vehicle ride comfort handling and stability andeconomic performance are not analyzed
For the MR damper the continuous variable character-istics of the damping coefficient can be achieved by con-trolling the magnetic field strength of the MR fluid [15ndash17] Anew kind of hybrid active suspension structure with ball screwactuator and MR damper is put forward Moreover in orderto ensure high response speed of the hybrid active suspensiona sky-hook algorithm which is easy to operate fast and robustin response is applied [18ndash20] And the semiactive modeand active mode of the hybrid suspension are analyzed underthe class B road surface At the same time the multimodecoordinated control system of the hybrid suspension isdesigned 0en the damping performance and energy con-sumption characteristics of the hybrid suspension under thecyclic driving condition are simulated and the bench tests ofthe hybrid suspension system are done
2 Structure and Principle of the HybridActive Suspension
0e structure of the hybrid active suspension system isshown in Figure 1 It is mainly composed of spring MRdamper ball screw actuator controller battery corre-sponding signal detection device and so on
0e controller performs semiactive control or activecontrol of the hybrid suspension system by detecting andjudging the relevant signals When the hybrid suspension issemiactively controlled the controller detects the sprung massacceleration in real time by the sprung mass accelerationsensor and the semiactive control force of the hybrid sus-pension is obtained through the sky-hook algorithm 0econtroller adjusts the controllable current to change themagnetic field strength of the MR fluid 0en the outputdamping force of the MR damper is changed and the sem-iactive control of the suspension is realized At the same timethe ball screw actuator transforms the up and down motion ofthe suspension into the rotation of the motor and then themotor generates electric energy which is stored in the battery
When the hybrid suspension is actively controlled thecontroller detects the sprung mass acceleration in real time bythe sprung mass acceleration sensor and the active controlforce of the hybrid suspension is obtained through the sky-
hook algorithm 0e controller adjusts the controllable currentto change the motor output torque 0e motor transforms therotation motion into the up and down motion and the activecontrol of the suspension is realized At this time the energyconsumed by the ball screw actuator is supplied by the batteryAnd when the hybrid suspension is actively controlled thecontroller detects the running speed of vehicle and controls theinput current of the MR damper according to different vehiclespeed values It makes the MR damper produce a dampingvalue that matches the vehicle speed In the semiactive andactive control of the suspension the energy consumed by theMR damper is all provided by the battery
3 Modeling of Hybrid Active SuspensionDynamic Model
31 Dynamic Model of Hybrid Active Suspension In thispaper a quarter vehicle dynamic model of the hybrid activesuspension is established and shown in Figure 2
Based on Newtonrsquos laws of motion the dynamic motionequations for the quarter vehicle suspension can beexpressed as
ms eurox2 + ks x2 minusx1 1113857 + cs _x2 minus _x1 1113857 F
mu eurox1 minus ks x2 minusx1 1113857minus cs _x2 minus _x1 1113857 + kt x1 minus z 1113857 minusF1113896
(1)
0e state variable and the output vector are selected asfollows
X x2 minusx1 _x2 x1 minus z _x11113858 1113859T
Y eurox2x2 minusx1kt x1 minus z 1113857 _x11113858 1113859T
(2)
where ms is sprung mass mu is unsprung mass ks is springstiffness coefficient F is control force of suspension (especiallyFb is semiactive control force and Fz is active control force) ktis tire stiffness coefficient z is displacement of road input x2 isdisplacement of sprung mass x1 is displacement of unsprungmass and cs is damping coefficient of MR damper
0en the state-space equations of suspension system canbe described as follows
_X AX + BU
Y CX +DU
⎧⎨
⎩ (3)
Sprung mass acceleration sensor
SpringBall screw actuator
Unsprung mass acceleration sensor
Sprung massSuspension controller
Battery
Vehicle speed sensor
Road surface Wheel
MR damperUnsprung mass
Figure 1 Structure of the hybrid active suspension
2 Shock and Vibration
where AΒC and D are the state matrix input matrixoutput matrix and transfer matrix respectively When thecontrol input force F is 0 it becomes passive suspension
A filtered white noise is adopted as the road surface inputmodel as follows
_z(t) minus2πf0z(t) + 2πG0u
1113968ω(t) (5)
where G0 is the coefficient of road irregularity f0 is lowercutoff frequency u is vehicle speed and ω(t) is unit whitenoise
0e simulation parameters of the hybrid suspension arelisted in Table 1
32 MR Damper Mathematical Model Ignoring the frictionand fluid inertia of the MR fluid the damping force model ofthe MR damper under mixed operation mode is given asfollows [21]
Fa minus24ηA2
pl
bh3 +2ηblh
1113888 1113889 _x2 minus _x1 1113857minus4lAp
η+ 2bl1113888 1113889τy (6)
where Fa is MR damper output damping force η isfluid dynamic viscosity l is the working plate length his the working plate distance τy is MR fluid shear stressAp is piston effective area and b is the working platewidth
According to formula (6) the damping force of the MRdamper includes the viscous damping force Fn which has afunction relationship with the piston speed of the MRdamper and coulomb damping force Fk which has afunction relationship with the control current of the MRdamper So formula (6) can be transformed into
Fa minusce _x2 minus _x1 1113857 + a1I2k + a2I
2k + a31113872 1113873sgn _x2 minus _x1 1113857 (7)
where ce a1 a2 and a3 are polynomial coefficients and Ik iscontrol current of the MR damper
From formulae (6)sim(7) because the viscous dampingforce Fn does not consume energy it can be equivalent to thedamping force produced by the traditional hydraulic shockabsorber 0e variable damping force of the hybrid activesuspension is the coulomb damping force Fk which isachieved by adjusting Ik
In the semiactive control of the hybrid suspension theinstantaneous energy consumption power and the energyconsumption of the MR damper are expressed as
Pb Fk middot_x2 minus _x1 1113857
ηb (8)
Wb 1113946t
0Pb dt (9)
where ηb is work efficiency of MR damper Pb is the MRdamper instantaneous energy consumption power of thehybrid suspension in semiactive control and Wb is the MRdamper energy consumption of the hybrid suspension insemiactive control
In the active control of the hybrid suspension the outputdamping force of the MR damper is expressed as
ms
mu
cs
F
z
ks
kt
x1
x2
Figure 2 Dynamic model of 2 DOF hybrid active suspension
Shock and Vibration 3
cs c0 + ck
c0 Fn
_x2 minus _x1( )
ck Fk
_x2 minus _x1( )
(10)
where c0 is viscous damping of the MR damper and ck isvariable damping of the MR damper
In the active control the instantaneous energy con-sumption power and the energy consumption of the MRdamper are expressed as
Pc cs minus c0( ) middot _x2 minus _x1( )2
ηb (11)
Wc intt
0Pc dt (12)
where Pc is the MR damper instantaneous energy con-sumption power of the hybrid suspension in active controlandWc is theMR damper energy consumption of the hybridsuspension in active control
e single rod MR damper is used in this paper and it isshown in Figure 3
By carrying out the characteristics test of the MR damperand analyzing the test data the relation diagrams of thedamping force-velocity curves of the MR damper are ob-tained and shown in Figure 4
33 Ball Screw Actuator Mathematical Model
331 e Characteristics of Ball Screw Actuator e ballscrew actuator is used not only to realize the active control ofthe hybrid suspension but also to recover the regenerativeenergy of the hybrid suspension the characteristics of whichhave a great inshyuence on the performance of the hybridsuspension
e characteristics of the ball screw actuator is mainlyaected by the back-EMF coecient ke and electric torquecoecient kT of the motor but the nonlinear characteristicsof the motor make ke and kT vary with the speed of themotor [22 23] It is necessary to gain the relationship be-tween ke kT and the motor speed by the motor test if aprecise mathematical model of the ball screw actuator isestablished
e prototype of the ball screw actuator is shown inFigure 5 It is made up of brushless DC motor ball screwupper and lower ears force sensor and so on
In this paper the peak value of counter electromotiveforce of the motor at dierent rotating speeds is measured bythe test At dierent rotating speeds the tting analyses of thepeak value are carried out and the result is shown in Figure 6
e peak of counter electromotive force of themotor andthe back-EMF coecient meet the following relationships
ke Vmax
22
radicn (13)
where Vmax is the peak of counter electromotive force and nis motor rotating speed
Table 1 Simulation parameters of the hybrid suspension
Parameters ValuesLower cuto frequency (Hz) 01Road irregularities coecient (m3) 64 times 10minus6
Sprung mass (kg) 38Unsprung mass (kg) 210Spring stiness coecient (kNm) 22Tire stiness coecient (kNm) 200
Upper ear
MR damper Connecting rod Lower ear
Force sensor
Figure 3 e MR damper prototype
ndash01 ndash005 0 005 01ndash600
ndash400
ndash200
0
200
400
600
Vibration speed (mmiddotsndash1)
Dam
ping
forc
e (N
)
02 A04 A06 A
08 A10 A
Figure 4 e damping force-velocity curves of the MR damper
Upper ear Force sensor
Ball screwBrushless DC motor Lower ear
Figure 5 e ball screw actuator
4 Shock and Vibration
e vibration velocity of the hybrid suspension and themotor rotating speed meet the following relationships
n ΔvLmiddot 60 (14)
Δv _x2 minus _x1 (15)
where Δv is the suspension vibration velocity and L is ballscrew lead
At dierent suspension vibration velocity the motorback-EMF coecient can be obtained from formulae(13)sim(15) and the tting curve of the motor back-EMFcoecient is shown in Figure 7
In Figure 7 the tting relationship between ke and Δv isexpressed as
ke 00299Δv3 minus 00542Δv2 + 00253Δv + 00049 (16)
en kT and ke meet the following relationships
kT 30πke (17)
e relationship between kT and Δv can be obtainedfrom formulae (16) and (17)
kT 02855Δv3 minus 05176Δv2 + 02416Δv + 00468 (18)
When testing the active output force of the ball screwactuator the ball screw actuator is powered by the same 72Vconstant voltage source as the battery pack terminal voltageBy adjusting PWM duty ratio the input voltage and currentof the motor are controlled by the controller and as a resultthe output force of the motor is gained e relationshipbetween the active output force of the ball screw actuatorand duty ratio is shown in Figure 8
From Figure 8 it can be seen that the ball screw actuatorhas good active output force characteristics and the non-linear relationship between the output force and the dutyratio is consistent with the nonlinear characteristics of themotor
332 Ball Screw Actuator Mathematical Model When themotor inductance is ignored in the active control of thehybrid suspension input voltage E and the output torqueTmof the motor meet the following relationships
E u + Izr (19)
Tm kT middot Iz (20)
u 60 _x2 minus _x1( )
Lmiddot ke (21)
Tm F middot L2π
(22)
ke kT middot π30
(23)
400 800 1200 1600 2000 2400 2800 3200 3600 40000
10
20
30
40
50
60
70
Rotating speed (rmin)
The p
eak
valu
e of c
ount
erel
ectr
omot
ive f
orce
(V)
The peak of counter electromotive forceTest fitting curve
Figure 6 e peak value curve of counter electromotive force ofthe motor
008 020 032 044 056 068 08000055
00060
00065
00070
00075
00080
00085
Suspension vibration speed (ms)
Back
-EM
F co
effic
ient
(Vr
middotmin
ndash1)
Test value data pointsTest fitting curve
Figure 7 e motor back-EMF coecient curve
ndash100 ndash50 0 50 100ndash600
ndash400
ndash200
0
200
400
600
Duty cycle ()
Act
uato
r act
ive o
utpu
t for
ce (N
)
Actuator active output forceTest fitting curve
Figure 8 Testing curve of the actuator active output force
Shock and Vibration 5
where u is induced electromotive force Iz is motor currentand r is internal resistance of the motor
In the active control of the hybrid suspension the in-stantaneous energy consumption power of the ball screwactuator can be expressed as
Pz E middot Iz
ηz (24)
0e instantaneous energy consumption power andconsumption energy can be obtained from formulae(19)sim(24) as follows
Pz _x2 minus _x1 1113857 middot F + FL2πkT 1113857
2middot r1113960 1113961
ηz (25)
Wz 1113946t
0Pz dt (26)
where Pz is motor instantaneous energy consumptionpower ηz is transfer efficiency of ball screw actuator andWzis motor energy consumption
In the energy regeneration of the hybrid suspension theelectromagnetic damping force generated by the ball screwactuator is expressed as
Fs minus2πkT
L1113888 1113889
2
middot_x2 minus _x1 1113857
R + rηz (27)
where Fs is the electromagnetic damping force and R isexternal resistance of the motor
0e instantaneous energy-regenerative power and regen-erative energy of the ball screw actuator can be expressed as
Pk 2πke
L1113888 1113889
2
middot_x2 minus _x1 1113857
2
R + rηz (28)
Wk 1113946t
0Pk dt (29)
where Pk is instantaneous energy-regenerative power andWk is regenerative energy
4 Multimode Coordination Control of HybridActive Suspension
41 -e Active Mode of Damping Switching Control 0esprung mass acceleration is the main evaluation index ofvehicle riding comfort and the dynamic tire load is closelyrelated to vehicle handling and stability An active controlmodel of the hybrid suspension is established by MATLABSimulink software to simulate and analyze the influence ofthe variable damping for the hybrid active suspension onvehicle riding comfort and handling and stability at differentvehicle speeds During the simulation the range of vehiclespeed v is 0ndash120 kmh and the vehicle speed is taken every10 kmh 0e variable damping range of the suspension is200ndash2000 Nmiddotsm and the value of the variable damping istaken every interval 100 Nmiddotsm 0e simulation time is 10 sand the value of r is 05 Ω 0e value of ηz is 097 and thevalue of ηb is 098 0e value of csky is 2000 Nmiddotsm
When the vehicle speed is 30 kmh and 100 kmh re-spectively the RMS of the sprung mass acceleration of thevehicle (aw) and the RMS of the dynamic tire load (DTLrms)change with the variable damping of the hybrid activesuspension as shown in Figures 9 and 10
From Figures 9 and 10 it can be seen that when thevehicle speed is 30 kmh the variable damping values thatmake aw of vehicle and DTLrms minimum are 400 Nmiddotsm and1000 Nmiddotsm respectively and when the vehicle speed is100 kmh the variable damping values that make aw ofvehicle and DTLrms minimum are 500 Nmiddotsm and 1100 Nmiddotsm respectively 0erefore at a certain vehicle speed thevariable damping of the hybrid active suspension cannotmake the best of the vehicle riding comfort and handling andstability at the same time
Suspension performance indexes include the sprung massacceleration suspension working space and dynamic tireload In this paper in order to balance vehicle riding comfortand handling and stability when choosing the variabledamping values of the hybrid suspension in the active controlfor the sprung mass acceleration suspension working spaceand dynamic tire load of the hybrid active suspension thequantitative normalizations and comparative analyses aredone 0at is at the same vehicle speed compared with thepassive suspension the improvement amplitudes of eachperformance index of the hybrid active suspension aremultiplied by different quantification factors and summedAnd the larger the sum the better the dynamic performanceof vehicle Among them the quantification factors of aw theRMS value of suspension working space (SWSrms) andDTLrms are 1 02703 and 01443 respectively [24 25] Andwhen the vehicle speed is 30 kmh and 100 kmh respectivelythe dynamic performance and the active control energyconsumption of the hybrid active suspension change with thevariable damping as shown in Figures 11 and 12
From Figures 11 and 12 when the vehicle speed is 30 kmh the variable damping value of the hybrid active suspensionis 800 Nmiddotsm which makes the vehicle dynamic performancethe best and the active energy consumption the least Whenthe vehicle speed is 100 kmh the variable damping value ofthe hybrid active suspension is 1000 Nmiddotsm which makes thevehicle dynamic performance the best and active energyconsumption the least 0erefore when the vehicle speed is30 kmh and 100 kmh respectively the optimal dampingvalues of the hybrid active suspension are 800 Nmiddotsm and1000 Nmiddotsm respectively When the vehicle speed is 0ndash120 kmh the optimal damping values of the hybrid activesuspension at different vehicle speeds are shown in Figure 13If the damping value which makes the vehicle dynamicperformance the best is different from the damping valuewhich makes the active control energy consumption the leastthe damping value which makes the vehicle dynamic per-formance the best is selected as the optimal damping value ofthe hybrid active suspension at the vehicle speed
When vehicle is in an accelerating or decelerating stateits speed changes rapidly and the range of change is wide sothe vehicle speed value is not easily detected in real time andin order to reduce the energy consumption of the hybridsuspension active controlled and to improve the vehicle
6 Shock and Vibration
riding comfort and handling and stability a variabledamping switching control strategy of the hybrid activesuspension is designed as follows
Fz minus csky middot _x2cs c0 + ck
_v 0
Fz minus csky middot _x2cs c0
_vne 0
(30)
where _v is vehicle acceleration and csky is sky-hook coecient
42 e Semiactive Mode of Feedback Adjustment of Elec-tromagnetic Damping Force MR damper can eectivelyperform semiactive control at ( _x2 minus _x1) _x2 gt 0 so the idealsemiactive control state for hybrid suspension is
Fb minus csky middot _x2 _x2 minus _x1( ) _x2 gt 0
0 _x2 minus _x1( ) _x2 le 0
(31)
From equation (31) the ideal semiactive control force ofthe hybrid suspension minus csky middot _x2 is only related to _x2 when
Vehicle dynamics performanceEnergy consumption of hybrid suspension
Figure 12 Relationship between the variable damping and vehicledynamic performance at 100 kmh speed
20 40 60 80 100 120750
800
850
900
950
1000
1050
Vehicle speed (kmh)
Shoc
k ab
sorb
er d
ampi
ngco
effic
ient
(Nmiddots
m)
Figure 13 Suspension optimal damping at dierent speeds
Shock and Vibration 7
the sky-hook coecient csky is constant However at thistime the ball screw actuator as a power feeding devicegenerates the electromagnetic damping force Fs and acts onthe suspension so that the actual semiactive control force ofthe suspension is dierent from the ideal semiactive controlforce minus csky middot _x2 In this paper the semiactive control modelof the hybrid suspension is established and the changeeects of the dierent output forces on the vehicle ridingcomfort and handling stability are analyzed by MATLABSimulink software e simulation speed is 70 kmh thesimulation time is 5 s and the value of R is 075 Ω edamping comparison of the hybrid suspension in semiactivecontrol is shown in Figure 14
Figure 14 shows that compared with the ideal semi-active control force the actual semiactive control force ofthe hybrid suspension shyuctuates violently and the abso-lute value of the actual semiactive control force is greaterthan the absolute value of the ideal semiactive control force|csky middot _x2| at certain times And a drastic change in theactual semiactive control force makes the suspension notreach ideal semiactive control eect Using electromag-netic damping force feedback adjustment to reduce thedierence between the ideal semiactive control forceminus csky middot _x2 and the actual semiactive control force themethod is as follows
When |csky middot _x2|gt |Fs| the semiactive control force of thehybrid suspension is provided by both the MR damper andthe ball screw actuator and at this point the controllerinputs a controllable current Ik to the MR damper so thatthe Fk output by the MR damper is minus csky middot _x2 minusFs Andwhen |csky middot _x2|le |Fs| the semiactive control force of thehybrid suspension is the Fs which is output by the ball screwactuator and at this point there is no controllable current Ikinput to theMR damper and the function of theMR damperis equivalent to a traditional shock absorber erefore thesemiactive control of the hybrid suspension does not havethe dead zone of traditional electromagnetic semiactivesuspension which helps to improve the semiactive controleect of the hybrid suspension
When there is feedback adjustment the semiactivecontrol force of the hybrid suspension is
Fb Fs Fs
∣∣∣∣∣∣∣∣ge csky middot _x2∣∣∣∣∣
∣∣∣∣∣
minus csky middot _x2 Fs∣∣∣∣∣∣∣∣lt csky middot _x2∣∣∣∣∣
∣∣∣∣∣
(32)
When there is feedback adjustment the Fk output by theMR damper is
Fk 0 Fs
∣∣∣∣∣∣∣∣ge csky middot _x2∣∣∣∣∣
∣∣∣∣∣
minus csky middot _x2 minusFs Fs∣∣∣∣∣∣∣∣lt csky middot _x2∣∣∣∣∣
∣∣∣∣∣
(33)
From equations (31)sim(33) when the electromagneticdamping force feedback adjustment is used the |Fk| outputby theMR damper decreases and when |Fk| decreases it canbe known from equations (8) and (9) that the energyconsumption of the MR damper decreases with it
e comparison of the semiactive control force of thehybrid suspension with or without the electromagneticdamping force feedback adjustment is shown in Figure 15
From Figure 15 the RMS of the ideal semiactive controlforce of the hybrid suspension is 3276 N and when there isno electromagnetic damping force feedback adjustmentthe RMS of the actual semiactive control force of thesuspension is 4041 N and the dierence between the actualsemiactive force of the suspension and the ideal semiactiveforce is 2335 When there is electromagnetic dampingforce feedback adjustment the RMS of the actual semi-active control force of the suspension is 3593 N and thedierence between the actual semiactive force of the sus-pension and the ideal semiactive force is 968 ereforewhen there is electromagnetic damping force feedbackadjustment the actual semiactive control force of thesuspension has a smaller shyuctuation amplitude whichhelps to improve the semiactive control eect of the hybridsuspension
e dynamic responses of the hybrid suspension with orwithout electromagnetic damping force feedback adjust-ment are shown in Figure 16 Among them the damper ofthe passive suspension is the original damper of the vehicleand its damping value is 1600Nmiddotsm
Table 2 shows the response RMS values of the hybridsuspension in semiactive control
From Table 2 compared with the passive suspensionwhen there is electromagnetic damping force feedbackadjustment aw SWSrms and DTLrms of the hybrid sus-pension are reduced by 1698 432 and 1068 re-spectively and compared with the nonfeedback semiactivecontrol when the feedback semiactive control is performedaw SWSrms and DTLrms of the hybrid suspension are re-duced by 252 863 and 671 respectively
From equations (9) and (29) the total system energy ofthe hybrid suspension in semiactive control is
W1 Wk minusWb (34)
whereW1 is the total system energy of the hybrid suspensionin semiactive control
0 1 2 3 4 5ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Hyb
rid su
spen
sion
forc
e (N
)
072 076 080
300
600
Ideal semiactive forceActual semiactive forceElectromagnetic damping force
Figure 14 Damping comparison of the hybrid suspension insemiactive control
8 Shock and Vibration
0 1 2 3 4 5ndash10
ndash5
0
5
10
15
Time (s)
Spru
ng m
ass a
ccel
erat
ion
(ms
2 )
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(a)
0 1 2 3 4 5ndash004
ndash002
000
002
004
006
008
Time (s)
Susp
ensio
n w
orki
ng sp
ace (
m)
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(b)
0 1 2 3 4 5ndash3000
ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Dyn
amic
tire
load
(N)
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(c)
Figure 16 e dynamic responses of the hybrid suspension in semiactive control (a) e response curves of sprung mass acceleration (b)e response curves of suspension working space (c) e response curves of dynamic tire load
0 1 2 3 4 5ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Hyb
rid su
spen
sion
forc
e (N
)
072 076 080
300
600
Ideal semiactive forceWithout feedback semiactive forceWith feedback semiactive force
Figure 15 Semiactive force of the hybrid suspension
Shock and Vibration 9
From equation (33) when the hybrid suspension sem-iactive controlled the curves of the total system energychange over time are shown in Figure 17
From Figure 17 when there is the nonfeedback semiactivecontrol the total system energy of the hybrid suspension is60 J And when there is the feedback semiactive control thetotal system energy of the hybrid suspension is 307 J
43 -e Design of Multimode Coordination ControllerWhen the hybrid suspension is actively controlled thesuspension has good vibration isolation performance buthigh energy consumption And when the hybrid suspensionis semiactively controlled the suspension has good eco-nomic performance but the control has limitations Con-sidering that the ball screw actuator can realize active controlof the hybrid suspension in any suspension state a multi-mode coordinated control strategy of the hybrid suspensionis designed
minus csky middot _x2 Fs1113868111386811138681113868
1113868111386811138681113868gt csky middot _x2
11138681113868111386811138681113868
11138681113868111386811138681113868
⎧⎪⎨
⎪⎩
cs c0
⎧⎪⎪⎪⎨
⎪⎪⎪⎩
_x2 minus _x1 1113857 _x2 gt 0
F Fz minus csky middot _x2
cs c0 + ck1113896 _x2 minus _x1 1113857 _x2 le 0
⎧⎪⎪⎪⎪⎪⎪⎪⎪⎪⎨
⎪⎪⎪⎪⎪⎪⎪⎪⎪⎩
(35)
0e frame diagram of the multimode coordinated controlstrategy of the hybrid suspension is shown in Figure 18
From equations (9) (12) (26) and (29) the total systemenergy of the hybrid suspension in multimode coordinatedcontrol is
W2 Wk minusWb minusWZ minusWc (36)
whereW2 is the total system energy of the hybrid suspensionin multimode coordinated control
A hybrid active suspension simulation model is estab-lished by using MATLABSimulink software From Fig-ure 11 when the hybrid suspension is actively controlled theenergy consumption of the MR damper is mainly affected bythe vehicle speed In order to verify the vibration isolationperformance and energy consumption performance of thehybrid active suspension in the cyclic driving conditionsthis article simulates vehicle urban and suburban conditionswhich is based on GBT 19233-2003 ldquoLight Vehicle FuelConsumption Test Methodrdquo among them the urban con-ditions include four cycle units each cycle time is 195 s thesuburban conditions include one cycle unit and the cycletime is 400 s [26] And the schematic diagram of the urbanconditions unit is shown in Figure 19
In order to simulate the vehicle acceleration signal avehicle speed variation model in different cycle units isestablished by using the signal builder function module inSimulink software On this basis the vehicle accelerationmodel in different cycle units is obtained and the vehicleacceleration model can be used as the switching controlmodel for the variable damping when the hybrid suspensionis actively controlled 0e shock absorbers of active sus-pension and passive suspension adopt the original damper ofvehicle and the damping value c1 is 1600 Nmiddotsm Howeverbecause of the long simulation time the dynamic responsecurves of the hybrid active suspension in urban and sub-urban conditions cannot be displayed Figure 20 shows thedynamic response curve of the hybrid active suspension inthe 145ndash175 s in the urban circulation unit among themvehicle in the 145ndash155 s is in a constant speed vehicle in the155ndash163 s is in the deceleration state and vehicle in the163ndash175 s is in other constant speed
0e dynamic response RMS values of the hybrid sus-pension in urban and suburban circulation units are shownin Tables 3 and 4 respectively
From Figure 20 and Tables 3 and 4 the dynamic re-sponses are good when the hybrid suspension is multimodecoordinated control When in the urban circulation unitcompared with the passive suspension aw SWSrms andDTLrms of the hybrid suspension are reduced by 3943225 and 2081 respectively Compared with the activesuspension aw and DTLrms of the hybrid suspension arereduced by 1322 and 840 respectively but SWSrms isincreased by 805 When in the suburban circulation unitcompared with the passive suspension aw SWSrms andDTLrms of the hybrid suspension are reduced by 3916314 and 1955 respectively Compared with the activesuspension aw and DTLrms of the hybrid suspension arereduced by 1199 and 732 respectively but SWSrms isincreased by 845
0e system energy of the hybrid suspension in urban andsuburban circulation units are shown in Figures 21 and 22respectively
Table 2 0e response root mean square values of the hybridsuspension in semiactive control
Without feedback semiactive controlWith feedback semiactive control
Figure 17 0e system energy of the hybrid suspension in semi-active control
10 Shock and Vibration
en the system energy values of the hybrid suspensionin urban and suburban circulation units are shown inTable 5
From Table 5 when in the urban circulation unit thesystem energy of the active suspension and hybrid sus-pension is minus2165 J and 38 J respectively And when in thesuburban circulation unit the system energy of the activesuspension and hybrid suspension is minus15071 J and minus122 Jrespectively e entire operation cycle includes 4 urbancycle units and 1 suburban cycle unit so the pure energy ofthe active suspension system during the entire operationcycle is minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid suspensionsystem basically realizes energy self-powered in theory
5 Test and Analysis
In order to verify the damping eect of the hybrid sus-pension system in active mode and semiactive mode ahybrid suspension vibration test system is designed as
Suspension controllerSpeed of sprung mass x2
Speed of unsprung mass x1Vehicle acceleration v
Vehicle suspension
Ball screw actuator
Output Fk
Active control force Fz
Ball screw actuator
Output Iz
Output Ik
Electromagnetic damping force Fs
Ideal semiactive control force FL
Output semiactive controlforce Fb = Fs
Output semiactive controlforce Fb = FL = Fk + Fs
x2 ndash x1
(x2 ndash x1) x2 gt 0
(x2 ndash x1) x2 le 0
Output Ik
MR damper
Adjust the duty cycle in real timeaccording to the relationshipbetween active output force
and duty cycle
No controllablecurrent output
MR damper
v ne 0
According to vand ∆v
v = 0
MR damper
Adjustable damping cs = c0 + ck
ndashcsky middot x2
ndashcsky middot x2
|Fs| ge |FL|
|Fs| lt |FL|
Figure 18 e frame diagram of the multimode coordinated control strategy of the hybrid suspension
Vehicle shifting
00
15
10
30
30
20
45
40
50
60
60
75 90 105 120Time (s)
Veh
icle
spee
d (k
mh
)
135 150 165 180 195
Figure 19 e schematic diagram of the urban conditions unit
Shock and Vibration 11
shown in Figure 23 During the test the MR damper has nocontrollable current input in the active control mode of thehybrid suspension and its damping value is always 800Nmiddotsm
e passive suspension semiactive suspension and activesuspension all adopt the original damper of the vehicle andthe damping value c1 is 1600Nmiddotsm
Figure 20e dynamic responses of the hybrid suspension (a)e response curves of sprungmass acceleration (b)e response curves ofsuspension working space (c) e response curves of dynamic tire load
12 Shock and Vibration
Because of the limitation of test conditions only thesprung mass acceleration dynamic response of the hybridsuspension is measured in this test e dynamic response ofthe sprung mass acceleration of the hybrid suspension inactive mode under random road is shown in Figure 24 epower spectrum of the sprung mass acceleration of thehybrid suspension in active mode is shown in Figure 25
Table 6 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in active mode
From Table 6 compared with passive suspension aw ofthe hybrid suspension in active mode is reduced by 3945and compared with active suspension aw of the hybridsuspension in active mode is reduced by 1432 And fromFigure 25 when the hybrid suspension is actively controlledcompared with passive suspension the vibration isolationperformance of the suspension in the low frequency andlow-frequency resonance areas is similar to that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withactive suspension the vibration isolation performance of thesuspension in the low frequency high frequency and res-onance regions is slightly better
e dynamic response of the sprung mass acceleration ofthe hybrid suspension in semiactive mode under randomroad is shown in Figure 26 e power spectrum of thesprung mass acceleration of the hybrid suspension insemiactive mode is shown in Figure 27
Table 7 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in semiactive mode
From Table 7 compared with passive suspension aw ofthe hybrid suspension in semiactive mode is reduced by1642 Compared with semiactive suspension aw of thehybrid suspension in active mode is reduced by 307 Andfrom Figure 27 when the hybrid suspension is semiactivelycontrolled compared with passive suspension the vibration
Table 3 e dynamic response root mean square values of thehybrid suspension in urban circulation unit
Figure 21 e system energy of the hybrid suspension in urbancirculation units
0 100 200 300 400ndash20000
ndash15000
ndash10000
ndash5000
0
5000
Time (s)
Susp
ensio
n sy
stem
ener
gy (J
)
Active suspensionHybrid suspension
Figure 22e system energy of the hybrid suspension in suburbancirculation units
Table 5 e system energy values of the hybrid suspension (J)
Cycle unit Active suspension Hybrid suspensionUrban minus2165 38Suburban minus15071 minus122
Figure 23 e test system of the hybrid active suspension
Shock and Vibration 13
isolation performance of the suspension in the low frequencyand low-frequency resonance areas is worse than that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withsemiactive suspension the vibration isolation performance ofthe suspension in the low frequency high frequency andresonance regions is slightly better
e energy consumption power and energy-regenerativepower of the hybrid suspension are shown in Figure 28
From Figure 28 in the active mode the average powerconsumption of the ball screw actuator is 2561W In thesemiactive mode the average regenerative power of the ballscrew actuator is 2696W and the average of theMR damperenergy consumption power is 178W erefore the pureaverage power consumption of the hybrid suspension isminus043W and the test results are basically consistent with thesimulation
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spr
ung
mas
s acc
eler
atio
n (m
s2 )
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 24 e dynamic response of the sprung mass accelerationof the hybrid suspension in active mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 25 e power spectrum of the sprung mass acceleration ofthe hybrid suspension in active mode
Table 6 e sprung mass acceleration RMS of the hybrid sus-pension in active mode
Indicators Passivesuspension
Activesuspension
Hybridsuspension
aw (ms2) 13378 09423 08074
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spru
ng m
ass a
ccel
erat
ion
(ms
2 )
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 26 e dynamic responses of the sprung mass accelerationof the hybrid suspension in semiactive mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 27e power spectrums of the sprungmass acceleration ofthe hybrid suspension in semiactive mode
Table 7 e sprung mass acceleration RMS of the hybrid sus-pension in semiactive mode
Indicators Passivesuspension
Semiactivesuspension
Hybridsuspension
aw (ms2) 13378 11536 11182
14 Shock and Vibration
6 Conclusion
(1) A ball screw actuator andMR damper are introducedinto vehicle suspension system and a new kind ofhybrid active suspension structure is put forwarde ball screw actuator is prototyped and thefunctional relationship between the back-EMF co-ecient the electromagnetic torque coecient of themotor and the suspension vibration speed is ob-tained by test analyses And the active output me-chanical properties of the ball screw actuator aretested and the results show that the actuator hasgood active output force characteristics
(2) e inshyuences of the variable damping value of thesuspension system on the riding comfort handlingand stability and energy consumption characteristicsof the hybrid suspension in the active control modeare analyzeden the optimal damping values of thehybrid suspension at dierent vehicle speeds aredesigned e eects of electromagnetic dampingforce on the actual semiactive force and the systemenergy of the suspension in the semiactive controlmode are analyzed and then the hybrid suspensionwith semiactive mode which has electromagneticdamping force feedback adjustment is designed Onthis basis a multimode coordinated control strategyfor the hybrid suspension is designed
(3) e damping performance and energy consumptioncharacteristics of the hybrid suspension under cyclicdriving condition are simulated by MATLABSimulink software and the results show that whenin the urban circulation unit compared with activesuspension aw and DTLrms of the hybrid suspensionare reduced by 1322 and 840 respectively Andwhen in the suburban circulation unit comparedwith the active suspension aw and DTLrms of thehybrid suspension are reduced by 1199 and 732respectively e pure energy of the active suspen-sion system during the entire operation cycle is
minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid sus-pension system basically realizes energy self-poweredin theory
(4) e eectiveness verication test of the hybrid sus-pension in active mode and semiactive mode controlis carried out and the results show that when thehybrid suspension is actively controlled comparedwith active suspension aw of the hybrid suspension inactive mode is reduced by 1432 When the hybridsuspension is semiactively controlled compared withsemiactive suspension aw of the hybrid suspension inactive mode is reduced by 307 e test and sim-ulation results are basically consistent and the testveries the correctness of the simulation
Data Availability
e data used to support the ndings of this study are in-cluded within the supplementary information les And thedata used to support the ndings of this study are availablefrom the corresponding author upon request
Conflicts of Interest
e authors declare that there are no conshyicts of interestregarding the publication of this paper
Acknowledgments
is work was supported by the National Natural ScienceFoundation of China (Grant no 51775426) Service LocalSpecial Program Support Project of Shaanxi ProvincialEducation Department (Grant no 17JF017) and XirsquoanScience and Technology Program Funding Project (Grantno 2017079CGRC042-XAKD007)
Supplementary Materials
e supplementary materials are this articlersquos experimentaldata including (1) MR damperrsquos characteristic test data (2)the motor counter electromotive forcersquos peak test data (3)the ball screw actuatorrsquos active output force test data (4) thehybrid suspensionrsquos sprung mass acceleration test data inactive mode (5) the hybrid suspensionrsquos sprung mass ac-celeration test data in semiactive mode and (6) the hybridsuspensionrsquos energy consumption power and energy re-generation power test data (Supplementary Materials)
References
[1] M Montazeri-Gh and O Kavianipour ldquoInvestigation of theactive electromagnetic suspension system considering hybridcontrol strategyrdquo Proceedings of the Institution of MechanicalEngineers Part C Journal of Mechanical Engineering Sciencevol 228 no 10 pp 1658ndash1669 2013
[2] H E Tseng and D Hrovat ldquoState of the art survey active andsemi-active suspension controlrdquo Vehicle System Dynamicsvol 53 no 7 pp 1034ndash1062 2015
0 1 2 3 4 50
20
40
60
80
Time (s)
Pow
er (W
)
Energy consumption power in active modeEnergy regenerative power in semiactive modeEnergy consumption power in semiactive mode
Figure 28 e energy consumption power and energy-re-generative power of the hybrid suspension
Shock and Vibration 15
[3] F Kou J Du Z Wang D Li and J Xu ldquoNonlinear modelingand coordinate optimization of a semi-active energy re-generative suspension with an electro-hydraulic actuatorrdquoChina Mechanical Engineering vol 28 no 14 pp 1701ndash17072017
[4] S H Zareh A Sarrafan A A A Khayyat and A ZabihollahldquoIntelligent semi-active vibration control of eleven degrees offreedom suspension system using magnetorheologicaldampersrdquo Journal of Mechanical Science and Technologyvol 26 no 3 pp 323ndash334 2012
[5] S A Chen X Li L J Zhao Y X Wang and Y B KimldquoDevelopment of a control method for an electromagneticsemi-active suspension reclaiming energy with varying chargevoltage in stepsrdquo International Journal of Automotive Tech-nology vol 16 no 5 pp 765ndash773 2015
[6] K Nakano Y Suda and S Nakadai ldquoSelf-powered activevibration control using a single electric actuatorrdquo Journal ofSound and Vibration vol 260 no 2 pp 213ndash235 2003
[7] K Nakano ldquoCombined type self-powered active vibrationcontrol of truck cabinsrdquo Vehicle System Dynamics vol 41no 6 pp 449ndash473 2004
[8] K Huang Y C Zhang F Yu and Y H Gu ldquoCoordinateoptimization for synthetical performance of electrical energy-regenerative active suspensionrdquo Journal of Shanghai JiaotongUniversity vol 43 no 2 pp 226ndash230 2009
[9] K Huang F Yu and Y C Zhang ldquoActive control of energy-regenerative electromagnetic suspension based on energy flowanalysisrdquo Journal of Shanghai Jiaotong University vol 45no 67 pp 1068ndash1073 2011
[10] D S Huang J Q Zhang Y L Liu L Yi and X Y WangldquoPerformance of a novel energy-regenerative active suspen-sion systemrdquo Journal of Chongqing University (English Edi-tion) vol 14 no 3 pp 109ndash118 2015
[11] B Ebrahimi H Bolandhemmat M B Khamesee andF Golnaraghi ldquoA hybrid electromagnetic shock absorber foractive vehicle suspension systemsrdquo Vehicle System Dynamicsvol 49 no 1-2 pp 311ndash332 2011
[12] S Tang L Chen R Wang X Sun and D Shi ldquoResearch onoptimal control of active suspension based on damping multi-modal switchingrdquo Journal of Guangxi University (NaturalScience) vol 39 no 2 pp 300ndash307 2014
[13] R Wang X Ma R Ding X Meng and L Chen ldquoResearchof multi-mode switching control system for hybrid sus-pension based on model referencerdquo Transactions of theChinese Society for Agricultural Machinery vol 48 no 7pp 353ndash360 2017
[14] R Wang Y Qian R Ding X Meng and J Xie ldquoDesign andtests for damping-stiffness of a hybrid electromagnetic sus-pension based on LQGrdquo Journal of Vibration and Shockvol 37 no 3 pp 61ndash65 2017
[15] B Vanavil K K Chaitanya and A S Rao ldquoImproved PIDcontroller design for unstable time delay processes based ondirect synthesis method and maximum sensitivityrdquo Taylorand Francis vol 46 no 8 pp 1349ndash1366 2015
[16] I Mihai and F Andronic ldquoBehavior of a semi-active sus-pension system versus a passive suspension system on anuneven road surfacerdquo Mechanics vol 20 no 1 pp 64ndash692014
[17] D Ngoduy ldquoLinear stability of a generalized multi-anticipative car following model with time delaysrdquo Com-munications in Nonlinear Science and Numerical Simulationvol 22 no 1 pp 420ndash426 2015
[18] H Zhang E Wang F Min R Subash and C Su ldquoSkyhook-based semi-active control of full-vehicle suspension with
magneto-rheological dampersrdquo Chinese Journal of Mechan-ical Engineering vol 26 no 3 pp 498ndash505 2013
[19] F Kou ldquoAn experimental study on the dynamic character-istics of vehicle semi-active seat suspension with magneto-rheological damperrdquo Automotive Engineering vol 37 no 11pp 1346ndash1352 2015
[20] Z Feng S Chen and Y Liang ldquoAn experimental study on thedynamic characteristics of a megneto-rheological semi-activesuspensionrdquoAutomotive Engineering vol 35 no 1 pp 72ndash772013
[21] Z Li and L Xu A New Type of Magnetorheological Damperand Semi-Active coNtrol Design-eory Science Press BeijingChina 2012
[22] B L J Gysen J J H Paulides J L G Janssen andE A Lomonova ldquoActive electromagnetic suspension systemfor improved vehicle dynamicsrdquo IEEE Transactions on Ve-hicular Technology vol 59 no 3 pp 1156ndash1163 2010
[23] Q N Wang S S Liu W H Wang and H Wei ldquoStructuredesign and parameter matching of ball-screw regenerativedamperrdquo Journal of Jilin University (Engineering and Tech-nology Edition) vol 42 no 5 pp 1100ndash1106 2012
[24] S Chen R He and S Lu ldquoEvaluating system of reclaimingenergy suspension comprehensive performancerdquo Trans-actions of the Chinese Society for Agricultural Machineryvol 37 no 7 pp 14ndash18 2006
[25] Y Fan Control Research on Vehicle Suspension with Electro-Hydrostatic Actuator Xirsquoan University of Science and Tech-nology Xirsquoan China 2017
[26] Z Yu Automobile -eory China Machine Press BeijingChina 2009
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Navigation and Observation
International Journal of
Hindawi
wwwhindawicom Volume 2018
Advances in
Multimedia
Submit your manuscripts atwwwhindawicom
where AΒC and D are the state matrix input matrixoutput matrix and transfer matrix respectively When thecontrol input force F is 0 it becomes passive suspension
A filtered white noise is adopted as the road surface inputmodel as follows
_z(t) minus2πf0z(t) + 2πG0u
1113968ω(t) (5)
where G0 is the coefficient of road irregularity f0 is lowercutoff frequency u is vehicle speed and ω(t) is unit whitenoise
0e simulation parameters of the hybrid suspension arelisted in Table 1
32 MR Damper Mathematical Model Ignoring the frictionand fluid inertia of the MR fluid the damping force model ofthe MR damper under mixed operation mode is given asfollows [21]
Fa minus24ηA2
pl
bh3 +2ηblh
1113888 1113889 _x2 minus _x1 1113857minus4lAp
η+ 2bl1113888 1113889τy (6)
where Fa is MR damper output damping force η isfluid dynamic viscosity l is the working plate length his the working plate distance τy is MR fluid shear stressAp is piston effective area and b is the working platewidth
According to formula (6) the damping force of the MRdamper includes the viscous damping force Fn which has afunction relationship with the piston speed of the MRdamper and coulomb damping force Fk which has afunction relationship with the control current of the MRdamper So formula (6) can be transformed into
Fa minusce _x2 minus _x1 1113857 + a1I2k + a2I
2k + a31113872 1113873sgn _x2 minus _x1 1113857 (7)
where ce a1 a2 and a3 are polynomial coefficients and Ik iscontrol current of the MR damper
From formulae (6)sim(7) because the viscous dampingforce Fn does not consume energy it can be equivalent to thedamping force produced by the traditional hydraulic shockabsorber 0e variable damping force of the hybrid activesuspension is the coulomb damping force Fk which isachieved by adjusting Ik
In the semiactive control of the hybrid suspension theinstantaneous energy consumption power and the energyconsumption of the MR damper are expressed as
Pb Fk middot_x2 minus _x1 1113857
ηb (8)
Wb 1113946t
0Pb dt (9)
where ηb is work efficiency of MR damper Pb is the MRdamper instantaneous energy consumption power of thehybrid suspension in semiactive control and Wb is the MRdamper energy consumption of the hybrid suspension insemiactive control
In the active control of the hybrid suspension the outputdamping force of the MR damper is expressed as
ms
mu
cs
F
z
ks
kt
x1
x2
Figure 2 Dynamic model of 2 DOF hybrid active suspension
Shock and Vibration 3
cs c0 + ck
c0 Fn
_x2 minus _x1( )
ck Fk
_x2 minus _x1( )
(10)
where c0 is viscous damping of the MR damper and ck isvariable damping of the MR damper
In the active control the instantaneous energy con-sumption power and the energy consumption of the MRdamper are expressed as
Pc cs minus c0( ) middot _x2 minus _x1( )2
ηb (11)
Wc intt
0Pc dt (12)
where Pc is the MR damper instantaneous energy con-sumption power of the hybrid suspension in active controlandWc is theMR damper energy consumption of the hybridsuspension in active control
e single rod MR damper is used in this paper and it isshown in Figure 3
By carrying out the characteristics test of the MR damperand analyzing the test data the relation diagrams of thedamping force-velocity curves of the MR damper are ob-tained and shown in Figure 4
33 Ball Screw Actuator Mathematical Model
331 e Characteristics of Ball Screw Actuator e ballscrew actuator is used not only to realize the active control ofthe hybrid suspension but also to recover the regenerativeenergy of the hybrid suspension the characteristics of whichhave a great inshyuence on the performance of the hybridsuspension
e characteristics of the ball screw actuator is mainlyaected by the back-EMF coecient ke and electric torquecoecient kT of the motor but the nonlinear characteristicsof the motor make ke and kT vary with the speed of themotor [22 23] It is necessary to gain the relationship be-tween ke kT and the motor speed by the motor test if aprecise mathematical model of the ball screw actuator isestablished
e prototype of the ball screw actuator is shown inFigure 5 It is made up of brushless DC motor ball screwupper and lower ears force sensor and so on
In this paper the peak value of counter electromotiveforce of the motor at dierent rotating speeds is measured bythe test At dierent rotating speeds the tting analyses of thepeak value are carried out and the result is shown in Figure 6
e peak of counter electromotive force of themotor andthe back-EMF coecient meet the following relationships
ke Vmax
22
radicn (13)
where Vmax is the peak of counter electromotive force and nis motor rotating speed
Table 1 Simulation parameters of the hybrid suspension
Parameters ValuesLower cuto frequency (Hz) 01Road irregularities coecient (m3) 64 times 10minus6
Sprung mass (kg) 38Unsprung mass (kg) 210Spring stiness coecient (kNm) 22Tire stiness coecient (kNm) 200
Upper ear
MR damper Connecting rod Lower ear
Force sensor
Figure 3 e MR damper prototype
ndash01 ndash005 0 005 01ndash600
ndash400
ndash200
0
200
400
600
Vibration speed (mmiddotsndash1)
Dam
ping
forc
e (N
)
02 A04 A06 A
08 A10 A
Figure 4 e damping force-velocity curves of the MR damper
Upper ear Force sensor
Ball screwBrushless DC motor Lower ear
Figure 5 e ball screw actuator
4 Shock and Vibration
e vibration velocity of the hybrid suspension and themotor rotating speed meet the following relationships
n ΔvLmiddot 60 (14)
Δv _x2 minus _x1 (15)
where Δv is the suspension vibration velocity and L is ballscrew lead
At dierent suspension vibration velocity the motorback-EMF coecient can be obtained from formulae(13)sim(15) and the tting curve of the motor back-EMFcoecient is shown in Figure 7
In Figure 7 the tting relationship between ke and Δv isexpressed as
ke 00299Δv3 minus 00542Δv2 + 00253Δv + 00049 (16)
en kT and ke meet the following relationships
kT 30πke (17)
e relationship between kT and Δv can be obtainedfrom formulae (16) and (17)
kT 02855Δv3 minus 05176Δv2 + 02416Δv + 00468 (18)
When testing the active output force of the ball screwactuator the ball screw actuator is powered by the same 72Vconstant voltage source as the battery pack terminal voltageBy adjusting PWM duty ratio the input voltage and currentof the motor are controlled by the controller and as a resultthe output force of the motor is gained e relationshipbetween the active output force of the ball screw actuatorand duty ratio is shown in Figure 8
From Figure 8 it can be seen that the ball screw actuatorhas good active output force characteristics and the non-linear relationship between the output force and the dutyratio is consistent with the nonlinear characteristics of themotor
332 Ball Screw Actuator Mathematical Model When themotor inductance is ignored in the active control of thehybrid suspension input voltage E and the output torqueTmof the motor meet the following relationships
E u + Izr (19)
Tm kT middot Iz (20)
u 60 _x2 minus _x1( )
Lmiddot ke (21)
Tm F middot L2π
(22)
ke kT middot π30
(23)
400 800 1200 1600 2000 2400 2800 3200 3600 40000
10
20
30
40
50
60
70
Rotating speed (rmin)
The p
eak
valu
e of c
ount
erel
ectr
omot
ive f
orce
(V)
The peak of counter electromotive forceTest fitting curve
Figure 6 e peak value curve of counter electromotive force ofthe motor
008 020 032 044 056 068 08000055
00060
00065
00070
00075
00080
00085
Suspension vibration speed (ms)
Back
-EM
F co
effic
ient
(Vr
middotmin
ndash1)
Test value data pointsTest fitting curve
Figure 7 e motor back-EMF coecient curve
ndash100 ndash50 0 50 100ndash600
ndash400
ndash200
0
200
400
600
Duty cycle ()
Act
uato
r act
ive o
utpu
t for
ce (N
)
Actuator active output forceTest fitting curve
Figure 8 Testing curve of the actuator active output force
Shock and Vibration 5
where u is induced electromotive force Iz is motor currentand r is internal resistance of the motor
In the active control of the hybrid suspension the in-stantaneous energy consumption power of the ball screwactuator can be expressed as
Pz E middot Iz
ηz (24)
0e instantaneous energy consumption power andconsumption energy can be obtained from formulae(19)sim(24) as follows
Pz _x2 minus _x1 1113857 middot F + FL2πkT 1113857
2middot r1113960 1113961
ηz (25)
Wz 1113946t
0Pz dt (26)
where Pz is motor instantaneous energy consumptionpower ηz is transfer efficiency of ball screw actuator andWzis motor energy consumption
In the energy regeneration of the hybrid suspension theelectromagnetic damping force generated by the ball screwactuator is expressed as
Fs minus2πkT
L1113888 1113889
2
middot_x2 minus _x1 1113857
R + rηz (27)
where Fs is the electromagnetic damping force and R isexternal resistance of the motor
0e instantaneous energy-regenerative power and regen-erative energy of the ball screw actuator can be expressed as
Pk 2πke
L1113888 1113889
2
middot_x2 minus _x1 1113857
2
R + rηz (28)
Wk 1113946t
0Pk dt (29)
where Pk is instantaneous energy-regenerative power andWk is regenerative energy
4 Multimode Coordination Control of HybridActive Suspension
41 -e Active Mode of Damping Switching Control 0esprung mass acceleration is the main evaluation index ofvehicle riding comfort and the dynamic tire load is closelyrelated to vehicle handling and stability An active controlmodel of the hybrid suspension is established by MATLABSimulink software to simulate and analyze the influence ofthe variable damping for the hybrid active suspension onvehicle riding comfort and handling and stability at differentvehicle speeds During the simulation the range of vehiclespeed v is 0ndash120 kmh and the vehicle speed is taken every10 kmh 0e variable damping range of the suspension is200ndash2000 Nmiddotsm and the value of the variable damping istaken every interval 100 Nmiddotsm 0e simulation time is 10 sand the value of r is 05 Ω 0e value of ηz is 097 and thevalue of ηb is 098 0e value of csky is 2000 Nmiddotsm
When the vehicle speed is 30 kmh and 100 kmh re-spectively the RMS of the sprung mass acceleration of thevehicle (aw) and the RMS of the dynamic tire load (DTLrms)change with the variable damping of the hybrid activesuspension as shown in Figures 9 and 10
From Figures 9 and 10 it can be seen that when thevehicle speed is 30 kmh the variable damping values thatmake aw of vehicle and DTLrms minimum are 400 Nmiddotsm and1000 Nmiddotsm respectively and when the vehicle speed is100 kmh the variable damping values that make aw ofvehicle and DTLrms minimum are 500 Nmiddotsm and 1100 Nmiddotsm respectively 0erefore at a certain vehicle speed thevariable damping of the hybrid active suspension cannotmake the best of the vehicle riding comfort and handling andstability at the same time
Suspension performance indexes include the sprung massacceleration suspension working space and dynamic tireload In this paper in order to balance vehicle riding comfortand handling and stability when choosing the variabledamping values of the hybrid suspension in the active controlfor the sprung mass acceleration suspension working spaceand dynamic tire load of the hybrid active suspension thequantitative normalizations and comparative analyses aredone 0at is at the same vehicle speed compared with thepassive suspension the improvement amplitudes of eachperformance index of the hybrid active suspension aremultiplied by different quantification factors and summedAnd the larger the sum the better the dynamic performanceof vehicle Among them the quantification factors of aw theRMS value of suspension working space (SWSrms) andDTLrms are 1 02703 and 01443 respectively [24 25] Andwhen the vehicle speed is 30 kmh and 100 kmh respectivelythe dynamic performance and the active control energyconsumption of the hybrid active suspension change with thevariable damping as shown in Figures 11 and 12
From Figures 11 and 12 when the vehicle speed is 30 kmh the variable damping value of the hybrid active suspensionis 800 Nmiddotsm which makes the vehicle dynamic performancethe best and the active energy consumption the least Whenthe vehicle speed is 100 kmh the variable damping value ofthe hybrid active suspension is 1000 Nmiddotsm which makes thevehicle dynamic performance the best and active energyconsumption the least 0erefore when the vehicle speed is30 kmh and 100 kmh respectively the optimal dampingvalues of the hybrid active suspension are 800 Nmiddotsm and1000 Nmiddotsm respectively When the vehicle speed is 0ndash120 kmh the optimal damping values of the hybrid activesuspension at different vehicle speeds are shown in Figure 13If the damping value which makes the vehicle dynamicperformance the best is different from the damping valuewhich makes the active control energy consumption the leastthe damping value which makes the vehicle dynamic per-formance the best is selected as the optimal damping value ofthe hybrid active suspension at the vehicle speed
When vehicle is in an accelerating or decelerating stateits speed changes rapidly and the range of change is wide sothe vehicle speed value is not easily detected in real time andin order to reduce the energy consumption of the hybridsuspension active controlled and to improve the vehicle
6 Shock and Vibration
riding comfort and handling and stability a variabledamping switching control strategy of the hybrid activesuspension is designed as follows
Fz minus csky middot _x2cs c0 + ck
_v 0
Fz minus csky middot _x2cs c0
_vne 0
(30)
where _v is vehicle acceleration and csky is sky-hook coecient
42 e Semiactive Mode of Feedback Adjustment of Elec-tromagnetic Damping Force MR damper can eectivelyperform semiactive control at ( _x2 minus _x1) _x2 gt 0 so the idealsemiactive control state for hybrid suspension is
Fb minus csky middot _x2 _x2 minus _x1( ) _x2 gt 0
0 _x2 minus _x1( ) _x2 le 0
(31)
From equation (31) the ideal semiactive control force ofthe hybrid suspension minus csky middot _x2 is only related to _x2 when
Vehicle dynamics performanceEnergy consumption of hybrid suspension
Figure 12 Relationship between the variable damping and vehicledynamic performance at 100 kmh speed
20 40 60 80 100 120750
800
850
900
950
1000
1050
Vehicle speed (kmh)
Shoc
k ab
sorb
er d
ampi
ngco
effic
ient
(Nmiddots
m)
Figure 13 Suspension optimal damping at dierent speeds
Shock and Vibration 7
the sky-hook coecient csky is constant However at thistime the ball screw actuator as a power feeding devicegenerates the electromagnetic damping force Fs and acts onthe suspension so that the actual semiactive control force ofthe suspension is dierent from the ideal semiactive controlforce minus csky middot _x2 In this paper the semiactive control modelof the hybrid suspension is established and the changeeects of the dierent output forces on the vehicle ridingcomfort and handling stability are analyzed by MATLABSimulink software e simulation speed is 70 kmh thesimulation time is 5 s and the value of R is 075 Ω edamping comparison of the hybrid suspension in semiactivecontrol is shown in Figure 14
Figure 14 shows that compared with the ideal semi-active control force the actual semiactive control force ofthe hybrid suspension shyuctuates violently and the abso-lute value of the actual semiactive control force is greaterthan the absolute value of the ideal semiactive control force|csky middot _x2| at certain times And a drastic change in theactual semiactive control force makes the suspension notreach ideal semiactive control eect Using electromag-netic damping force feedback adjustment to reduce thedierence between the ideal semiactive control forceminus csky middot _x2 and the actual semiactive control force themethod is as follows
When |csky middot _x2|gt |Fs| the semiactive control force of thehybrid suspension is provided by both the MR damper andthe ball screw actuator and at this point the controllerinputs a controllable current Ik to the MR damper so thatthe Fk output by the MR damper is minus csky middot _x2 minusFs Andwhen |csky middot _x2|le |Fs| the semiactive control force of thehybrid suspension is the Fs which is output by the ball screwactuator and at this point there is no controllable current Ikinput to theMR damper and the function of theMR damperis equivalent to a traditional shock absorber erefore thesemiactive control of the hybrid suspension does not havethe dead zone of traditional electromagnetic semiactivesuspension which helps to improve the semiactive controleect of the hybrid suspension
When there is feedback adjustment the semiactivecontrol force of the hybrid suspension is
Fb Fs Fs
∣∣∣∣∣∣∣∣ge csky middot _x2∣∣∣∣∣
∣∣∣∣∣
minus csky middot _x2 Fs∣∣∣∣∣∣∣∣lt csky middot _x2∣∣∣∣∣
∣∣∣∣∣
(32)
When there is feedback adjustment the Fk output by theMR damper is
Fk 0 Fs
∣∣∣∣∣∣∣∣ge csky middot _x2∣∣∣∣∣
∣∣∣∣∣
minus csky middot _x2 minusFs Fs∣∣∣∣∣∣∣∣lt csky middot _x2∣∣∣∣∣
∣∣∣∣∣
(33)
From equations (31)sim(33) when the electromagneticdamping force feedback adjustment is used the |Fk| outputby theMR damper decreases and when |Fk| decreases it canbe known from equations (8) and (9) that the energyconsumption of the MR damper decreases with it
e comparison of the semiactive control force of thehybrid suspension with or without the electromagneticdamping force feedback adjustment is shown in Figure 15
From Figure 15 the RMS of the ideal semiactive controlforce of the hybrid suspension is 3276 N and when there isno electromagnetic damping force feedback adjustmentthe RMS of the actual semiactive control force of thesuspension is 4041 N and the dierence between the actualsemiactive force of the suspension and the ideal semiactiveforce is 2335 When there is electromagnetic dampingforce feedback adjustment the RMS of the actual semi-active control force of the suspension is 3593 N and thedierence between the actual semiactive force of the sus-pension and the ideal semiactive force is 968 ereforewhen there is electromagnetic damping force feedbackadjustment the actual semiactive control force of thesuspension has a smaller shyuctuation amplitude whichhelps to improve the semiactive control eect of the hybridsuspension
e dynamic responses of the hybrid suspension with orwithout electromagnetic damping force feedback adjust-ment are shown in Figure 16 Among them the damper ofthe passive suspension is the original damper of the vehicleand its damping value is 1600Nmiddotsm
Table 2 shows the response RMS values of the hybridsuspension in semiactive control
From Table 2 compared with the passive suspensionwhen there is electromagnetic damping force feedbackadjustment aw SWSrms and DTLrms of the hybrid sus-pension are reduced by 1698 432 and 1068 re-spectively and compared with the nonfeedback semiactivecontrol when the feedback semiactive control is performedaw SWSrms and DTLrms of the hybrid suspension are re-duced by 252 863 and 671 respectively
From equations (9) and (29) the total system energy ofthe hybrid suspension in semiactive control is
W1 Wk minusWb (34)
whereW1 is the total system energy of the hybrid suspensionin semiactive control
0 1 2 3 4 5ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Hyb
rid su
spen
sion
forc
e (N
)
072 076 080
300
600
Ideal semiactive forceActual semiactive forceElectromagnetic damping force
Figure 14 Damping comparison of the hybrid suspension insemiactive control
8 Shock and Vibration
0 1 2 3 4 5ndash10
ndash5
0
5
10
15
Time (s)
Spru
ng m
ass a
ccel
erat
ion
(ms
2 )
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(a)
0 1 2 3 4 5ndash004
ndash002
000
002
004
006
008
Time (s)
Susp
ensio
n w
orki
ng sp
ace (
m)
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(b)
0 1 2 3 4 5ndash3000
ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Dyn
amic
tire
load
(N)
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(c)
Figure 16 e dynamic responses of the hybrid suspension in semiactive control (a) e response curves of sprung mass acceleration (b)e response curves of suspension working space (c) e response curves of dynamic tire load
0 1 2 3 4 5ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Hyb
rid su
spen
sion
forc
e (N
)
072 076 080
300
600
Ideal semiactive forceWithout feedback semiactive forceWith feedback semiactive force
Figure 15 Semiactive force of the hybrid suspension
Shock and Vibration 9
From equation (33) when the hybrid suspension sem-iactive controlled the curves of the total system energychange over time are shown in Figure 17
From Figure 17 when there is the nonfeedback semiactivecontrol the total system energy of the hybrid suspension is60 J And when there is the feedback semiactive control thetotal system energy of the hybrid suspension is 307 J
43 -e Design of Multimode Coordination ControllerWhen the hybrid suspension is actively controlled thesuspension has good vibration isolation performance buthigh energy consumption And when the hybrid suspensionis semiactively controlled the suspension has good eco-nomic performance but the control has limitations Con-sidering that the ball screw actuator can realize active controlof the hybrid suspension in any suspension state a multi-mode coordinated control strategy of the hybrid suspensionis designed
minus csky middot _x2 Fs1113868111386811138681113868
1113868111386811138681113868gt csky middot _x2
11138681113868111386811138681113868
11138681113868111386811138681113868
⎧⎪⎨
⎪⎩
cs c0
⎧⎪⎪⎪⎨
⎪⎪⎪⎩
_x2 minus _x1 1113857 _x2 gt 0
F Fz minus csky middot _x2
cs c0 + ck1113896 _x2 minus _x1 1113857 _x2 le 0
⎧⎪⎪⎪⎪⎪⎪⎪⎪⎪⎨
⎪⎪⎪⎪⎪⎪⎪⎪⎪⎩
(35)
0e frame diagram of the multimode coordinated controlstrategy of the hybrid suspension is shown in Figure 18
From equations (9) (12) (26) and (29) the total systemenergy of the hybrid suspension in multimode coordinatedcontrol is
W2 Wk minusWb minusWZ minusWc (36)
whereW2 is the total system energy of the hybrid suspensionin multimode coordinated control
A hybrid active suspension simulation model is estab-lished by using MATLABSimulink software From Fig-ure 11 when the hybrid suspension is actively controlled theenergy consumption of the MR damper is mainly affected bythe vehicle speed In order to verify the vibration isolationperformance and energy consumption performance of thehybrid active suspension in the cyclic driving conditionsthis article simulates vehicle urban and suburban conditionswhich is based on GBT 19233-2003 ldquoLight Vehicle FuelConsumption Test Methodrdquo among them the urban con-ditions include four cycle units each cycle time is 195 s thesuburban conditions include one cycle unit and the cycletime is 400 s [26] And the schematic diagram of the urbanconditions unit is shown in Figure 19
In order to simulate the vehicle acceleration signal avehicle speed variation model in different cycle units isestablished by using the signal builder function module inSimulink software On this basis the vehicle accelerationmodel in different cycle units is obtained and the vehicleacceleration model can be used as the switching controlmodel for the variable damping when the hybrid suspensionis actively controlled 0e shock absorbers of active sus-pension and passive suspension adopt the original damper ofvehicle and the damping value c1 is 1600 Nmiddotsm Howeverbecause of the long simulation time the dynamic responsecurves of the hybrid active suspension in urban and sub-urban conditions cannot be displayed Figure 20 shows thedynamic response curve of the hybrid active suspension inthe 145ndash175 s in the urban circulation unit among themvehicle in the 145ndash155 s is in a constant speed vehicle in the155ndash163 s is in the deceleration state and vehicle in the163ndash175 s is in other constant speed
0e dynamic response RMS values of the hybrid sus-pension in urban and suburban circulation units are shownin Tables 3 and 4 respectively
From Figure 20 and Tables 3 and 4 the dynamic re-sponses are good when the hybrid suspension is multimodecoordinated control When in the urban circulation unitcompared with the passive suspension aw SWSrms andDTLrms of the hybrid suspension are reduced by 3943225 and 2081 respectively Compared with the activesuspension aw and DTLrms of the hybrid suspension arereduced by 1322 and 840 respectively but SWSrms isincreased by 805 When in the suburban circulation unitcompared with the passive suspension aw SWSrms andDTLrms of the hybrid suspension are reduced by 3916314 and 1955 respectively Compared with the activesuspension aw and DTLrms of the hybrid suspension arereduced by 1199 and 732 respectively but SWSrms isincreased by 845
0e system energy of the hybrid suspension in urban andsuburban circulation units are shown in Figures 21 and 22respectively
Table 2 0e response root mean square values of the hybridsuspension in semiactive control
Without feedback semiactive controlWith feedback semiactive control
Figure 17 0e system energy of the hybrid suspension in semi-active control
10 Shock and Vibration
en the system energy values of the hybrid suspensionin urban and suburban circulation units are shown inTable 5
From Table 5 when in the urban circulation unit thesystem energy of the active suspension and hybrid sus-pension is minus2165 J and 38 J respectively And when in thesuburban circulation unit the system energy of the activesuspension and hybrid suspension is minus15071 J and minus122 Jrespectively e entire operation cycle includes 4 urbancycle units and 1 suburban cycle unit so the pure energy ofthe active suspension system during the entire operationcycle is minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid suspensionsystem basically realizes energy self-powered in theory
5 Test and Analysis
In order to verify the damping eect of the hybrid sus-pension system in active mode and semiactive mode ahybrid suspension vibration test system is designed as
Suspension controllerSpeed of sprung mass x2
Speed of unsprung mass x1Vehicle acceleration v
Vehicle suspension
Ball screw actuator
Output Fk
Active control force Fz
Ball screw actuator
Output Iz
Output Ik
Electromagnetic damping force Fs
Ideal semiactive control force FL
Output semiactive controlforce Fb = Fs
Output semiactive controlforce Fb = FL = Fk + Fs
x2 ndash x1
(x2 ndash x1) x2 gt 0
(x2 ndash x1) x2 le 0
Output Ik
MR damper
Adjust the duty cycle in real timeaccording to the relationshipbetween active output force
and duty cycle
No controllablecurrent output
MR damper
v ne 0
According to vand ∆v
v = 0
MR damper
Adjustable damping cs = c0 + ck
ndashcsky middot x2
ndashcsky middot x2
|Fs| ge |FL|
|Fs| lt |FL|
Figure 18 e frame diagram of the multimode coordinated control strategy of the hybrid suspension
Vehicle shifting
00
15
10
30
30
20
45
40
50
60
60
75 90 105 120Time (s)
Veh
icle
spee
d (k
mh
)
135 150 165 180 195
Figure 19 e schematic diagram of the urban conditions unit
Shock and Vibration 11
shown in Figure 23 During the test the MR damper has nocontrollable current input in the active control mode of thehybrid suspension and its damping value is always 800Nmiddotsm
e passive suspension semiactive suspension and activesuspension all adopt the original damper of the vehicle andthe damping value c1 is 1600Nmiddotsm
Figure 20e dynamic responses of the hybrid suspension (a)e response curves of sprungmass acceleration (b)e response curves ofsuspension working space (c) e response curves of dynamic tire load
12 Shock and Vibration
Because of the limitation of test conditions only thesprung mass acceleration dynamic response of the hybridsuspension is measured in this test e dynamic response ofthe sprung mass acceleration of the hybrid suspension inactive mode under random road is shown in Figure 24 epower spectrum of the sprung mass acceleration of thehybrid suspension in active mode is shown in Figure 25
Table 6 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in active mode
From Table 6 compared with passive suspension aw ofthe hybrid suspension in active mode is reduced by 3945and compared with active suspension aw of the hybridsuspension in active mode is reduced by 1432 And fromFigure 25 when the hybrid suspension is actively controlledcompared with passive suspension the vibration isolationperformance of the suspension in the low frequency andlow-frequency resonance areas is similar to that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withactive suspension the vibration isolation performance of thesuspension in the low frequency high frequency and res-onance regions is slightly better
e dynamic response of the sprung mass acceleration ofthe hybrid suspension in semiactive mode under randomroad is shown in Figure 26 e power spectrum of thesprung mass acceleration of the hybrid suspension insemiactive mode is shown in Figure 27
Table 7 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in semiactive mode
From Table 7 compared with passive suspension aw ofthe hybrid suspension in semiactive mode is reduced by1642 Compared with semiactive suspension aw of thehybrid suspension in active mode is reduced by 307 Andfrom Figure 27 when the hybrid suspension is semiactivelycontrolled compared with passive suspension the vibration
Table 3 e dynamic response root mean square values of thehybrid suspension in urban circulation unit
Figure 21 e system energy of the hybrid suspension in urbancirculation units
0 100 200 300 400ndash20000
ndash15000
ndash10000
ndash5000
0
5000
Time (s)
Susp
ensio
n sy
stem
ener
gy (J
)
Active suspensionHybrid suspension
Figure 22e system energy of the hybrid suspension in suburbancirculation units
Table 5 e system energy values of the hybrid suspension (J)
Cycle unit Active suspension Hybrid suspensionUrban minus2165 38Suburban minus15071 minus122
Figure 23 e test system of the hybrid active suspension
Shock and Vibration 13
isolation performance of the suspension in the low frequencyand low-frequency resonance areas is worse than that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withsemiactive suspension the vibration isolation performance ofthe suspension in the low frequency high frequency andresonance regions is slightly better
e energy consumption power and energy-regenerativepower of the hybrid suspension are shown in Figure 28
From Figure 28 in the active mode the average powerconsumption of the ball screw actuator is 2561W In thesemiactive mode the average regenerative power of the ballscrew actuator is 2696W and the average of theMR damperenergy consumption power is 178W erefore the pureaverage power consumption of the hybrid suspension isminus043W and the test results are basically consistent with thesimulation
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spr
ung
mas
s acc
eler
atio
n (m
s2 )
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 24 e dynamic response of the sprung mass accelerationof the hybrid suspension in active mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 25 e power spectrum of the sprung mass acceleration ofthe hybrid suspension in active mode
Table 6 e sprung mass acceleration RMS of the hybrid sus-pension in active mode
Indicators Passivesuspension
Activesuspension
Hybridsuspension
aw (ms2) 13378 09423 08074
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spru
ng m
ass a
ccel
erat
ion
(ms
2 )
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 26 e dynamic responses of the sprung mass accelerationof the hybrid suspension in semiactive mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 27e power spectrums of the sprungmass acceleration ofthe hybrid suspension in semiactive mode
Table 7 e sprung mass acceleration RMS of the hybrid sus-pension in semiactive mode
Indicators Passivesuspension
Semiactivesuspension
Hybridsuspension
aw (ms2) 13378 11536 11182
14 Shock and Vibration
6 Conclusion
(1) A ball screw actuator andMR damper are introducedinto vehicle suspension system and a new kind ofhybrid active suspension structure is put forwarde ball screw actuator is prototyped and thefunctional relationship between the back-EMF co-ecient the electromagnetic torque coecient of themotor and the suspension vibration speed is ob-tained by test analyses And the active output me-chanical properties of the ball screw actuator aretested and the results show that the actuator hasgood active output force characteristics
(2) e inshyuences of the variable damping value of thesuspension system on the riding comfort handlingand stability and energy consumption characteristicsof the hybrid suspension in the active control modeare analyzeden the optimal damping values of thehybrid suspension at dierent vehicle speeds aredesigned e eects of electromagnetic dampingforce on the actual semiactive force and the systemenergy of the suspension in the semiactive controlmode are analyzed and then the hybrid suspensionwith semiactive mode which has electromagneticdamping force feedback adjustment is designed Onthis basis a multimode coordinated control strategyfor the hybrid suspension is designed
(3) e damping performance and energy consumptioncharacteristics of the hybrid suspension under cyclicdriving condition are simulated by MATLABSimulink software and the results show that whenin the urban circulation unit compared with activesuspension aw and DTLrms of the hybrid suspensionare reduced by 1322 and 840 respectively Andwhen in the suburban circulation unit comparedwith the active suspension aw and DTLrms of thehybrid suspension are reduced by 1199 and 732respectively e pure energy of the active suspen-sion system during the entire operation cycle is
minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid sus-pension system basically realizes energy self-poweredin theory
(4) e eectiveness verication test of the hybrid sus-pension in active mode and semiactive mode controlis carried out and the results show that when thehybrid suspension is actively controlled comparedwith active suspension aw of the hybrid suspension inactive mode is reduced by 1432 When the hybridsuspension is semiactively controlled compared withsemiactive suspension aw of the hybrid suspension inactive mode is reduced by 307 e test and sim-ulation results are basically consistent and the testveries the correctness of the simulation
Data Availability
e data used to support the ndings of this study are in-cluded within the supplementary information les And thedata used to support the ndings of this study are availablefrom the corresponding author upon request
Conflicts of Interest
e authors declare that there are no conshyicts of interestregarding the publication of this paper
Acknowledgments
is work was supported by the National Natural ScienceFoundation of China (Grant no 51775426) Service LocalSpecial Program Support Project of Shaanxi ProvincialEducation Department (Grant no 17JF017) and XirsquoanScience and Technology Program Funding Project (Grantno 2017079CGRC042-XAKD007)
Supplementary Materials
e supplementary materials are this articlersquos experimentaldata including (1) MR damperrsquos characteristic test data (2)the motor counter electromotive forcersquos peak test data (3)the ball screw actuatorrsquos active output force test data (4) thehybrid suspensionrsquos sprung mass acceleration test data inactive mode (5) the hybrid suspensionrsquos sprung mass ac-celeration test data in semiactive mode and (6) the hybridsuspensionrsquos energy consumption power and energy re-generation power test data (Supplementary Materials)
References
[1] M Montazeri-Gh and O Kavianipour ldquoInvestigation of theactive electromagnetic suspension system considering hybridcontrol strategyrdquo Proceedings of the Institution of MechanicalEngineers Part C Journal of Mechanical Engineering Sciencevol 228 no 10 pp 1658ndash1669 2013
[2] H E Tseng and D Hrovat ldquoState of the art survey active andsemi-active suspension controlrdquo Vehicle System Dynamicsvol 53 no 7 pp 1034ndash1062 2015
0 1 2 3 4 50
20
40
60
80
Time (s)
Pow
er (W
)
Energy consumption power in active modeEnergy regenerative power in semiactive modeEnergy consumption power in semiactive mode
Figure 28 e energy consumption power and energy-re-generative power of the hybrid suspension
Shock and Vibration 15
[3] F Kou J Du Z Wang D Li and J Xu ldquoNonlinear modelingand coordinate optimization of a semi-active energy re-generative suspension with an electro-hydraulic actuatorrdquoChina Mechanical Engineering vol 28 no 14 pp 1701ndash17072017
[4] S H Zareh A Sarrafan A A A Khayyat and A ZabihollahldquoIntelligent semi-active vibration control of eleven degrees offreedom suspension system using magnetorheologicaldampersrdquo Journal of Mechanical Science and Technologyvol 26 no 3 pp 323ndash334 2012
[5] S A Chen X Li L J Zhao Y X Wang and Y B KimldquoDevelopment of a control method for an electromagneticsemi-active suspension reclaiming energy with varying chargevoltage in stepsrdquo International Journal of Automotive Tech-nology vol 16 no 5 pp 765ndash773 2015
[6] K Nakano Y Suda and S Nakadai ldquoSelf-powered activevibration control using a single electric actuatorrdquo Journal ofSound and Vibration vol 260 no 2 pp 213ndash235 2003
[7] K Nakano ldquoCombined type self-powered active vibrationcontrol of truck cabinsrdquo Vehicle System Dynamics vol 41no 6 pp 449ndash473 2004
[8] K Huang Y C Zhang F Yu and Y H Gu ldquoCoordinateoptimization for synthetical performance of electrical energy-regenerative active suspensionrdquo Journal of Shanghai JiaotongUniversity vol 43 no 2 pp 226ndash230 2009
[9] K Huang F Yu and Y C Zhang ldquoActive control of energy-regenerative electromagnetic suspension based on energy flowanalysisrdquo Journal of Shanghai Jiaotong University vol 45no 67 pp 1068ndash1073 2011
[10] D S Huang J Q Zhang Y L Liu L Yi and X Y WangldquoPerformance of a novel energy-regenerative active suspen-sion systemrdquo Journal of Chongqing University (English Edi-tion) vol 14 no 3 pp 109ndash118 2015
[11] B Ebrahimi H Bolandhemmat M B Khamesee andF Golnaraghi ldquoA hybrid electromagnetic shock absorber foractive vehicle suspension systemsrdquo Vehicle System Dynamicsvol 49 no 1-2 pp 311ndash332 2011
[12] S Tang L Chen R Wang X Sun and D Shi ldquoResearch onoptimal control of active suspension based on damping multi-modal switchingrdquo Journal of Guangxi University (NaturalScience) vol 39 no 2 pp 300ndash307 2014
[13] R Wang X Ma R Ding X Meng and L Chen ldquoResearchof multi-mode switching control system for hybrid sus-pension based on model referencerdquo Transactions of theChinese Society for Agricultural Machinery vol 48 no 7pp 353ndash360 2017
[14] R Wang Y Qian R Ding X Meng and J Xie ldquoDesign andtests for damping-stiffness of a hybrid electromagnetic sus-pension based on LQGrdquo Journal of Vibration and Shockvol 37 no 3 pp 61ndash65 2017
[15] B Vanavil K K Chaitanya and A S Rao ldquoImproved PIDcontroller design for unstable time delay processes based ondirect synthesis method and maximum sensitivityrdquo Taylorand Francis vol 46 no 8 pp 1349ndash1366 2015
[16] I Mihai and F Andronic ldquoBehavior of a semi-active sus-pension system versus a passive suspension system on anuneven road surfacerdquo Mechanics vol 20 no 1 pp 64ndash692014
[17] D Ngoduy ldquoLinear stability of a generalized multi-anticipative car following model with time delaysrdquo Com-munications in Nonlinear Science and Numerical Simulationvol 22 no 1 pp 420ndash426 2015
[18] H Zhang E Wang F Min R Subash and C Su ldquoSkyhook-based semi-active control of full-vehicle suspension with
magneto-rheological dampersrdquo Chinese Journal of Mechan-ical Engineering vol 26 no 3 pp 498ndash505 2013
[19] F Kou ldquoAn experimental study on the dynamic character-istics of vehicle semi-active seat suspension with magneto-rheological damperrdquo Automotive Engineering vol 37 no 11pp 1346ndash1352 2015
[20] Z Feng S Chen and Y Liang ldquoAn experimental study on thedynamic characteristics of a megneto-rheological semi-activesuspensionrdquoAutomotive Engineering vol 35 no 1 pp 72ndash772013
[21] Z Li and L Xu A New Type of Magnetorheological Damperand Semi-Active coNtrol Design-eory Science Press BeijingChina 2012
[22] B L J Gysen J J H Paulides J L G Janssen andE A Lomonova ldquoActive electromagnetic suspension systemfor improved vehicle dynamicsrdquo IEEE Transactions on Ve-hicular Technology vol 59 no 3 pp 1156ndash1163 2010
[23] Q N Wang S S Liu W H Wang and H Wei ldquoStructuredesign and parameter matching of ball-screw regenerativedamperrdquo Journal of Jilin University (Engineering and Tech-nology Edition) vol 42 no 5 pp 1100ndash1106 2012
[24] S Chen R He and S Lu ldquoEvaluating system of reclaimingenergy suspension comprehensive performancerdquo Trans-actions of the Chinese Society for Agricultural Machineryvol 37 no 7 pp 14ndash18 2006
[25] Y Fan Control Research on Vehicle Suspension with Electro-Hydrostatic Actuator Xirsquoan University of Science and Tech-nology Xirsquoan China 2017
[26] Z Yu Automobile -eory China Machine Press BeijingChina 2009
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Navigation and Observation
International Journal of
Hindawi
wwwhindawicom Volume 2018
Advances in
Multimedia
Submit your manuscripts atwwwhindawicom
cs c0 + ck
c0 Fn
_x2 minus _x1( )
ck Fk
_x2 minus _x1( )
(10)
where c0 is viscous damping of the MR damper and ck isvariable damping of the MR damper
In the active control the instantaneous energy con-sumption power and the energy consumption of the MRdamper are expressed as
Pc cs minus c0( ) middot _x2 minus _x1( )2
ηb (11)
Wc intt
0Pc dt (12)
where Pc is the MR damper instantaneous energy con-sumption power of the hybrid suspension in active controlandWc is theMR damper energy consumption of the hybridsuspension in active control
e single rod MR damper is used in this paper and it isshown in Figure 3
By carrying out the characteristics test of the MR damperand analyzing the test data the relation diagrams of thedamping force-velocity curves of the MR damper are ob-tained and shown in Figure 4
33 Ball Screw Actuator Mathematical Model
331 e Characteristics of Ball Screw Actuator e ballscrew actuator is used not only to realize the active control ofthe hybrid suspension but also to recover the regenerativeenergy of the hybrid suspension the characteristics of whichhave a great inshyuence on the performance of the hybridsuspension
e characteristics of the ball screw actuator is mainlyaected by the back-EMF coecient ke and electric torquecoecient kT of the motor but the nonlinear characteristicsof the motor make ke and kT vary with the speed of themotor [22 23] It is necessary to gain the relationship be-tween ke kT and the motor speed by the motor test if aprecise mathematical model of the ball screw actuator isestablished
e prototype of the ball screw actuator is shown inFigure 5 It is made up of brushless DC motor ball screwupper and lower ears force sensor and so on
In this paper the peak value of counter electromotiveforce of the motor at dierent rotating speeds is measured bythe test At dierent rotating speeds the tting analyses of thepeak value are carried out and the result is shown in Figure 6
e peak of counter electromotive force of themotor andthe back-EMF coecient meet the following relationships
ke Vmax
22
radicn (13)
where Vmax is the peak of counter electromotive force and nis motor rotating speed
Table 1 Simulation parameters of the hybrid suspension
Parameters ValuesLower cuto frequency (Hz) 01Road irregularities coecient (m3) 64 times 10minus6
Sprung mass (kg) 38Unsprung mass (kg) 210Spring stiness coecient (kNm) 22Tire stiness coecient (kNm) 200
Upper ear
MR damper Connecting rod Lower ear
Force sensor
Figure 3 e MR damper prototype
ndash01 ndash005 0 005 01ndash600
ndash400
ndash200
0
200
400
600
Vibration speed (mmiddotsndash1)
Dam
ping
forc
e (N
)
02 A04 A06 A
08 A10 A
Figure 4 e damping force-velocity curves of the MR damper
Upper ear Force sensor
Ball screwBrushless DC motor Lower ear
Figure 5 e ball screw actuator
4 Shock and Vibration
e vibration velocity of the hybrid suspension and themotor rotating speed meet the following relationships
n ΔvLmiddot 60 (14)
Δv _x2 minus _x1 (15)
where Δv is the suspension vibration velocity and L is ballscrew lead
At dierent suspension vibration velocity the motorback-EMF coecient can be obtained from formulae(13)sim(15) and the tting curve of the motor back-EMFcoecient is shown in Figure 7
In Figure 7 the tting relationship between ke and Δv isexpressed as
ke 00299Δv3 minus 00542Δv2 + 00253Δv + 00049 (16)
en kT and ke meet the following relationships
kT 30πke (17)
e relationship between kT and Δv can be obtainedfrom formulae (16) and (17)
kT 02855Δv3 minus 05176Δv2 + 02416Δv + 00468 (18)
When testing the active output force of the ball screwactuator the ball screw actuator is powered by the same 72Vconstant voltage source as the battery pack terminal voltageBy adjusting PWM duty ratio the input voltage and currentof the motor are controlled by the controller and as a resultthe output force of the motor is gained e relationshipbetween the active output force of the ball screw actuatorand duty ratio is shown in Figure 8
From Figure 8 it can be seen that the ball screw actuatorhas good active output force characteristics and the non-linear relationship between the output force and the dutyratio is consistent with the nonlinear characteristics of themotor
332 Ball Screw Actuator Mathematical Model When themotor inductance is ignored in the active control of thehybrid suspension input voltage E and the output torqueTmof the motor meet the following relationships
E u + Izr (19)
Tm kT middot Iz (20)
u 60 _x2 minus _x1( )
Lmiddot ke (21)
Tm F middot L2π
(22)
ke kT middot π30
(23)
400 800 1200 1600 2000 2400 2800 3200 3600 40000
10
20
30
40
50
60
70
Rotating speed (rmin)
The p
eak
valu
e of c
ount
erel
ectr
omot
ive f
orce
(V)
The peak of counter electromotive forceTest fitting curve
Figure 6 e peak value curve of counter electromotive force ofthe motor
008 020 032 044 056 068 08000055
00060
00065
00070
00075
00080
00085
Suspension vibration speed (ms)
Back
-EM
F co
effic
ient
(Vr
middotmin
ndash1)
Test value data pointsTest fitting curve
Figure 7 e motor back-EMF coecient curve
ndash100 ndash50 0 50 100ndash600
ndash400
ndash200
0
200
400
600
Duty cycle ()
Act
uato
r act
ive o
utpu
t for
ce (N
)
Actuator active output forceTest fitting curve
Figure 8 Testing curve of the actuator active output force
Shock and Vibration 5
where u is induced electromotive force Iz is motor currentand r is internal resistance of the motor
In the active control of the hybrid suspension the in-stantaneous energy consumption power of the ball screwactuator can be expressed as
Pz E middot Iz
ηz (24)
0e instantaneous energy consumption power andconsumption energy can be obtained from formulae(19)sim(24) as follows
Pz _x2 minus _x1 1113857 middot F + FL2πkT 1113857
2middot r1113960 1113961
ηz (25)
Wz 1113946t
0Pz dt (26)
where Pz is motor instantaneous energy consumptionpower ηz is transfer efficiency of ball screw actuator andWzis motor energy consumption
In the energy regeneration of the hybrid suspension theelectromagnetic damping force generated by the ball screwactuator is expressed as
Fs minus2πkT
L1113888 1113889
2
middot_x2 minus _x1 1113857
R + rηz (27)
where Fs is the electromagnetic damping force and R isexternal resistance of the motor
0e instantaneous energy-regenerative power and regen-erative energy of the ball screw actuator can be expressed as
Pk 2πke
L1113888 1113889
2
middot_x2 minus _x1 1113857
2
R + rηz (28)
Wk 1113946t
0Pk dt (29)
where Pk is instantaneous energy-regenerative power andWk is regenerative energy
4 Multimode Coordination Control of HybridActive Suspension
41 -e Active Mode of Damping Switching Control 0esprung mass acceleration is the main evaluation index ofvehicle riding comfort and the dynamic tire load is closelyrelated to vehicle handling and stability An active controlmodel of the hybrid suspension is established by MATLABSimulink software to simulate and analyze the influence ofthe variable damping for the hybrid active suspension onvehicle riding comfort and handling and stability at differentvehicle speeds During the simulation the range of vehiclespeed v is 0ndash120 kmh and the vehicle speed is taken every10 kmh 0e variable damping range of the suspension is200ndash2000 Nmiddotsm and the value of the variable damping istaken every interval 100 Nmiddotsm 0e simulation time is 10 sand the value of r is 05 Ω 0e value of ηz is 097 and thevalue of ηb is 098 0e value of csky is 2000 Nmiddotsm
When the vehicle speed is 30 kmh and 100 kmh re-spectively the RMS of the sprung mass acceleration of thevehicle (aw) and the RMS of the dynamic tire load (DTLrms)change with the variable damping of the hybrid activesuspension as shown in Figures 9 and 10
From Figures 9 and 10 it can be seen that when thevehicle speed is 30 kmh the variable damping values thatmake aw of vehicle and DTLrms minimum are 400 Nmiddotsm and1000 Nmiddotsm respectively and when the vehicle speed is100 kmh the variable damping values that make aw ofvehicle and DTLrms minimum are 500 Nmiddotsm and 1100 Nmiddotsm respectively 0erefore at a certain vehicle speed thevariable damping of the hybrid active suspension cannotmake the best of the vehicle riding comfort and handling andstability at the same time
Suspension performance indexes include the sprung massacceleration suspension working space and dynamic tireload In this paper in order to balance vehicle riding comfortand handling and stability when choosing the variabledamping values of the hybrid suspension in the active controlfor the sprung mass acceleration suspension working spaceand dynamic tire load of the hybrid active suspension thequantitative normalizations and comparative analyses aredone 0at is at the same vehicle speed compared with thepassive suspension the improvement amplitudes of eachperformance index of the hybrid active suspension aremultiplied by different quantification factors and summedAnd the larger the sum the better the dynamic performanceof vehicle Among them the quantification factors of aw theRMS value of suspension working space (SWSrms) andDTLrms are 1 02703 and 01443 respectively [24 25] Andwhen the vehicle speed is 30 kmh and 100 kmh respectivelythe dynamic performance and the active control energyconsumption of the hybrid active suspension change with thevariable damping as shown in Figures 11 and 12
From Figures 11 and 12 when the vehicle speed is 30 kmh the variable damping value of the hybrid active suspensionis 800 Nmiddotsm which makes the vehicle dynamic performancethe best and the active energy consumption the least Whenthe vehicle speed is 100 kmh the variable damping value ofthe hybrid active suspension is 1000 Nmiddotsm which makes thevehicle dynamic performance the best and active energyconsumption the least 0erefore when the vehicle speed is30 kmh and 100 kmh respectively the optimal dampingvalues of the hybrid active suspension are 800 Nmiddotsm and1000 Nmiddotsm respectively When the vehicle speed is 0ndash120 kmh the optimal damping values of the hybrid activesuspension at different vehicle speeds are shown in Figure 13If the damping value which makes the vehicle dynamicperformance the best is different from the damping valuewhich makes the active control energy consumption the leastthe damping value which makes the vehicle dynamic per-formance the best is selected as the optimal damping value ofthe hybrid active suspension at the vehicle speed
When vehicle is in an accelerating or decelerating stateits speed changes rapidly and the range of change is wide sothe vehicle speed value is not easily detected in real time andin order to reduce the energy consumption of the hybridsuspension active controlled and to improve the vehicle
6 Shock and Vibration
riding comfort and handling and stability a variabledamping switching control strategy of the hybrid activesuspension is designed as follows
Fz minus csky middot _x2cs c0 + ck
_v 0
Fz minus csky middot _x2cs c0
_vne 0
(30)
where _v is vehicle acceleration and csky is sky-hook coecient
42 e Semiactive Mode of Feedback Adjustment of Elec-tromagnetic Damping Force MR damper can eectivelyperform semiactive control at ( _x2 minus _x1) _x2 gt 0 so the idealsemiactive control state for hybrid suspension is
Fb minus csky middot _x2 _x2 minus _x1( ) _x2 gt 0
0 _x2 minus _x1( ) _x2 le 0
(31)
From equation (31) the ideal semiactive control force ofthe hybrid suspension minus csky middot _x2 is only related to _x2 when
Vehicle dynamics performanceEnergy consumption of hybrid suspension
Figure 12 Relationship between the variable damping and vehicledynamic performance at 100 kmh speed
20 40 60 80 100 120750
800
850
900
950
1000
1050
Vehicle speed (kmh)
Shoc
k ab
sorb
er d
ampi
ngco
effic
ient
(Nmiddots
m)
Figure 13 Suspension optimal damping at dierent speeds
Shock and Vibration 7
the sky-hook coecient csky is constant However at thistime the ball screw actuator as a power feeding devicegenerates the electromagnetic damping force Fs and acts onthe suspension so that the actual semiactive control force ofthe suspension is dierent from the ideal semiactive controlforce minus csky middot _x2 In this paper the semiactive control modelof the hybrid suspension is established and the changeeects of the dierent output forces on the vehicle ridingcomfort and handling stability are analyzed by MATLABSimulink software e simulation speed is 70 kmh thesimulation time is 5 s and the value of R is 075 Ω edamping comparison of the hybrid suspension in semiactivecontrol is shown in Figure 14
Figure 14 shows that compared with the ideal semi-active control force the actual semiactive control force ofthe hybrid suspension shyuctuates violently and the abso-lute value of the actual semiactive control force is greaterthan the absolute value of the ideal semiactive control force|csky middot _x2| at certain times And a drastic change in theactual semiactive control force makes the suspension notreach ideal semiactive control eect Using electromag-netic damping force feedback adjustment to reduce thedierence between the ideal semiactive control forceminus csky middot _x2 and the actual semiactive control force themethod is as follows
When |csky middot _x2|gt |Fs| the semiactive control force of thehybrid suspension is provided by both the MR damper andthe ball screw actuator and at this point the controllerinputs a controllable current Ik to the MR damper so thatthe Fk output by the MR damper is minus csky middot _x2 minusFs Andwhen |csky middot _x2|le |Fs| the semiactive control force of thehybrid suspension is the Fs which is output by the ball screwactuator and at this point there is no controllable current Ikinput to theMR damper and the function of theMR damperis equivalent to a traditional shock absorber erefore thesemiactive control of the hybrid suspension does not havethe dead zone of traditional electromagnetic semiactivesuspension which helps to improve the semiactive controleect of the hybrid suspension
When there is feedback adjustment the semiactivecontrol force of the hybrid suspension is
Fb Fs Fs
∣∣∣∣∣∣∣∣ge csky middot _x2∣∣∣∣∣
∣∣∣∣∣
minus csky middot _x2 Fs∣∣∣∣∣∣∣∣lt csky middot _x2∣∣∣∣∣
∣∣∣∣∣
(32)
When there is feedback adjustment the Fk output by theMR damper is
Fk 0 Fs
∣∣∣∣∣∣∣∣ge csky middot _x2∣∣∣∣∣
∣∣∣∣∣
minus csky middot _x2 minusFs Fs∣∣∣∣∣∣∣∣lt csky middot _x2∣∣∣∣∣
∣∣∣∣∣
(33)
From equations (31)sim(33) when the electromagneticdamping force feedback adjustment is used the |Fk| outputby theMR damper decreases and when |Fk| decreases it canbe known from equations (8) and (9) that the energyconsumption of the MR damper decreases with it
e comparison of the semiactive control force of thehybrid suspension with or without the electromagneticdamping force feedback adjustment is shown in Figure 15
From Figure 15 the RMS of the ideal semiactive controlforce of the hybrid suspension is 3276 N and when there isno electromagnetic damping force feedback adjustmentthe RMS of the actual semiactive control force of thesuspension is 4041 N and the dierence between the actualsemiactive force of the suspension and the ideal semiactiveforce is 2335 When there is electromagnetic dampingforce feedback adjustment the RMS of the actual semi-active control force of the suspension is 3593 N and thedierence between the actual semiactive force of the sus-pension and the ideal semiactive force is 968 ereforewhen there is electromagnetic damping force feedbackadjustment the actual semiactive control force of thesuspension has a smaller shyuctuation amplitude whichhelps to improve the semiactive control eect of the hybridsuspension
e dynamic responses of the hybrid suspension with orwithout electromagnetic damping force feedback adjust-ment are shown in Figure 16 Among them the damper ofthe passive suspension is the original damper of the vehicleand its damping value is 1600Nmiddotsm
Table 2 shows the response RMS values of the hybridsuspension in semiactive control
From Table 2 compared with the passive suspensionwhen there is electromagnetic damping force feedbackadjustment aw SWSrms and DTLrms of the hybrid sus-pension are reduced by 1698 432 and 1068 re-spectively and compared with the nonfeedback semiactivecontrol when the feedback semiactive control is performedaw SWSrms and DTLrms of the hybrid suspension are re-duced by 252 863 and 671 respectively
From equations (9) and (29) the total system energy ofthe hybrid suspension in semiactive control is
W1 Wk minusWb (34)
whereW1 is the total system energy of the hybrid suspensionin semiactive control
0 1 2 3 4 5ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Hyb
rid su
spen
sion
forc
e (N
)
072 076 080
300
600
Ideal semiactive forceActual semiactive forceElectromagnetic damping force
Figure 14 Damping comparison of the hybrid suspension insemiactive control
8 Shock and Vibration
0 1 2 3 4 5ndash10
ndash5
0
5
10
15
Time (s)
Spru
ng m
ass a
ccel
erat
ion
(ms
2 )
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(a)
0 1 2 3 4 5ndash004
ndash002
000
002
004
006
008
Time (s)
Susp
ensio
n w
orki
ng sp
ace (
m)
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(b)
0 1 2 3 4 5ndash3000
ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Dyn
amic
tire
load
(N)
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(c)
Figure 16 e dynamic responses of the hybrid suspension in semiactive control (a) e response curves of sprung mass acceleration (b)e response curves of suspension working space (c) e response curves of dynamic tire load
0 1 2 3 4 5ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Hyb
rid su
spen
sion
forc
e (N
)
072 076 080
300
600
Ideal semiactive forceWithout feedback semiactive forceWith feedback semiactive force
Figure 15 Semiactive force of the hybrid suspension
Shock and Vibration 9
From equation (33) when the hybrid suspension sem-iactive controlled the curves of the total system energychange over time are shown in Figure 17
From Figure 17 when there is the nonfeedback semiactivecontrol the total system energy of the hybrid suspension is60 J And when there is the feedback semiactive control thetotal system energy of the hybrid suspension is 307 J
43 -e Design of Multimode Coordination ControllerWhen the hybrid suspension is actively controlled thesuspension has good vibration isolation performance buthigh energy consumption And when the hybrid suspensionis semiactively controlled the suspension has good eco-nomic performance but the control has limitations Con-sidering that the ball screw actuator can realize active controlof the hybrid suspension in any suspension state a multi-mode coordinated control strategy of the hybrid suspensionis designed
minus csky middot _x2 Fs1113868111386811138681113868
1113868111386811138681113868gt csky middot _x2
11138681113868111386811138681113868
11138681113868111386811138681113868
⎧⎪⎨
⎪⎩
cs c0
⎧⎪⎪⎪⎨
⎪⎪⎪⎩
_x2 minus _x1 1113857 _x2 gt 0
F Fz minus csky middot _x2
cs c0 + ck1113896 _x2 minus _x1 1113857 _x2 le 0
⎧⎪⎪⎪⎪⎪⎪⎪⎪⎪⎨
⎪⎪⎪⎪⎪⎪⎪⎪⎪⎩
(35)
0e frame diagram of the multimode coordinated controlstrategy of the hybrid suspension is shown in Figure 18
From equations (9) (12) (26) and (29) the total systemenergy of the hybrid suspension in multimode coordinatedcontrol is
W2 Wk minusWb minusWZ minusWc (36)
whereW2 is the total system energy of the hybrid suspensionin multimode coordinated control
A hybrid active suspension simulation model is estab-lished by using MATLABSimulink software From Fig-ure 11 when the hybrid suspension is actively controlled theenergy consumption of the MR damper is mainly affected bythe vehicle speed In order to verify the vibration isolationperformance and energy consumption performance of thehybrid active suspension in the cyclic driving conditionsthis article simulates vehicle urban and suburban conditionswhich is based on GBT 19233-2003 ldquoLight Vehicle FuelConsumption Test Methodrdquo among them the urban con-ditions include four cycle units each cycle time is 195 s thesuburban conditions include one cycle unit and the cycletime is 400 s [26] And the schematic diagram of the urbanconditions unit is shown in Figure 19
In order to simulate the vehicle acceleration signal avehicle speed variation model in different cycle units isestablished by using the signal builder function module inSimulink software On this basis the vehicle accelerationmodel in different cycle units is obtained and the vehicleacceleration model can be used as the switching controlmodel for the variable damping when the hybrid suspensionis actively controlled 0e shock absorbers of active sus-pension and passive suspension adopt the original damper ofvehicle and the damping value c1 is 1600 Nmiddotsm Howeverbecause of the long simulation time the dynamic responsecurves of the hybrid active suspension in urban and sub-urban conditions cannot be displayed Figure 20 shows thedynamic response curve of the hybrid active suspension inthe 145ndash175 s in the urban circulation unit among themvehicle in the 145ndash155 s is in a constant speed vehicle in the155ndash163 s is in the deceleration state and vehicle in the163ndash175 s is in other constant speed
0e dynamic response RMS values of the hybrid sus-pension in urban and suburban circulation units are shownin Tables 3 and 4 respectively
From Figure 20 and Tables 3 and 4 the dynamic re-sponses are good when the hybrid suspension is multimodecoordinated control When in the urban circulation unitcompared with the passive suspension aw SWSrms andDTLrms of the hybrid suspension are reduced by 3943225 and 2081 respectively Compared with the activesuspension aw and DTLrms of the hybrid suspension arereduced by 1322 and 840 respectively but SWSrms isincreased by 805 When in the suburban circulation unitcompared with the passive suspension aw SWSrms andDTLrms of the hybrid suspension are reduced by 3916314 and 1955 respectively Compared with the activesuspension aw and DTLrms of the hybrid suspension arereduced by 1199 and 732 respectively but SWSrms isincreased by 845
0e system energy of the hybrid suspension in urban andsuburban circulation units are shown in Figures 21 and 22respectively
Table 2 0e response root mean square values of the hybridsuspension in semiactive control
Without feedback semiactive controlWith feedback semiactive control
Figure 17 0e system energy of the hybrid suspension in semi-active control
10 Shock and Vibration
en the system energy values of the hybrid suspensionin urban and suburban circulation units are shown inTable 5
From Table 5 when in the urban circulation unit thesystem energy of the active suspension and hybrid sus-pension is minus2165 J and 38 J respectively And when in thesuburban circulation unit the system energy of the activesuspension and hybrid suspension is minus15071 J and minus122 Jrespectively e entire operation cycle includes 4 urbancycle units and 1 suburban cycle unit so the pure energy ofthe active suspension system during the entire operationcycle is minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid suspensionsystem basically realizes energy self-powered in theory
5 Test and Analysis
In order to verify the damping eect of the hybrid sus-pension system in active mode and semiactive mode ahybrid suspension vibration test system is designed as
Suspension controllerSpeed of sprung mass x2
Speed of unsprung mass x1Vehicle acceleration v
Vehicle suspension
Ball screw actuator
Output Fk
Active control force Fz
Ball screw actuator
Output Iz
Output Ik
Electromagnetic damping force Fs
Ideal semiactive control force FL
Output semiactive controlforce Fb = Fs
Output semiactive controlforce Fb = FL = Fk + Fs
x2 ndash x1
(x2 ndash x1) x2 gt 0
(x2 ndash x1) x2 le 0
Output Ik
MR damper
Adjust the duty cycle in real timeaccording to the relationshipbetween active output force
and duty cycle
No controllablecurrent output
MR damper
v ne 0
According to vand ∆v
v = 0
MR damper
Adjustable damping cs = c0 + ck
ndashcsky middot x2
ndashcsky middot x2
|Fs| ge |FL|
|Fs| lt |FL|
Figure 18 e frame diagram of the multimode coordinated control strategy of the hybrid suspension
Vehicle shifting
00
15
10
30
30
20
45
40
50
60
60
75 90 105 120Time (s)
Veh
icle
spee
d (k
mh
)
135 150 165 180 195
Figure 19 e schematic diagram of the urban conditions unit
Shock and Vibration 11
shown in Figure 23 During the test the MR damper has nocontrollable current input in the active control mode of thehybrid suspension and its damping value is always 800Nmiddotsm
e passive suspension semiactive suspension and activesuspension all adopt the original damper of the vehicle andthe damping value c1 is 1600Nmiddotsm
Figure 20e dynamic responses of the hybrid suspension (a)e response curves of sprungmass acceleration (b)e response curves ofsuspension working space (c) e response curves of dynamic tire load
12 Shock and Vibration
Because of the limitation of test conditions only thesprung mass acceleration dynamic response of the hybridsuspension is measured in this test e dynamic response ofthe sprung mass acceleration of the hybrid suspension inactive mode under random road is shown in Figure 24 epower spectrum of the sprung mass acceleration of thehybrid suspension in active mode is shown in Figure 25
Table 6 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in active mode
From Table 6 compared with passive suspension aw ofthe hybrid suspension in active mode is reduced by 3945and compared with active suspension aw of the hybridsuspension in active mode is reduced by 1432 And fromFigure 25 when the hybrid suspension is actively controlledcompared with passive suspension the vibration isolationperformance of the suspension in the low frequency andlow-frequency resonance areas is similar to that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withactive suspension the vibration isolation performance of thesuspension in the low frequency high frequency and res-onance regions is slightly better
e dynamic response of the sprung mass acceleration ofthe hybrid suspension in semiactive mode under randomroad is shown in Figure 26 e power spectrum of thesprung mass acceleration of the hybrid suspension insemiactive mode is shown in Figure 27
Table 7 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in semiactive mode
From Table 7 compared with passive suspension aw ofthe hybrid suspension in semiactive mode is reduced by1642 Compared with semiactive suspension aw of thehybrid suspension in active mode is reduced by 307 Andfrom Figure 27 when the hybrid suspension is semiactivelycontrolled compared with passive suspension the vibration
Table 3 e dynamic response root mean square values of thehybrid suspension in urban circulation unit
Figure 21 e system energy of the hybrid suspension in urbancirculation units
0 100 200 300 400ndash20000
ndash15000
ndash10000
ndash5000
0
5000
Time (s)
Susp
ensio
n sy
stem
ener
gy (J
)
Active suspensionHybrid suspension
Figure 22e system energy of the hybrid suspension in suburbancirculation units
Table 5 e system energy values of the hybrid suspension (J)
Cycle unit Active suspension Hybrid suspensionUrban minus2165 38Suburban minus15071 minus122
Figure 23 e test system of the hybrid active suspension
Shock and Vibration 13
isolation performance of the suspension in the low frequencyand low-frequency resonance areas is worse than that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withsemiactive suspension the vibration isolation performance ofthe suspension in the low frequency high frequency andresonance regions is slightly better
e energy consumption power and energy-regenerativepower of the hybrid suspension are shown in Figure 28
From Figure 28 in the active mode the average powerconsumption of the ball screw actuator is 2561W In thesemiactive mode the average regenerative power of the ballscrew actuator is 2696W and the average of theMR damperenergy consumption power is 178W erefore the pureaverage power consumption of the hybrid suspension isminus043W and the test results are basically consistent with thesimulation
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spr
ung
mas
s acc
eler
atio
n (m
s2 )
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 24 e dynamic response of the sprung mass accelerationof the hybrid suspension in active mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 25 e power spectrum of the sprung mass acceleration ofthe hybrid suspension in active mode
Table 6 e sprung mass acceleration RMS of the hybrid sus-pension in active mode
Indicators Passivesuspension
Activesuspension
Hybridsuspension
aw (ms2) 13378 09423 08074
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spru
ng m
ass a
ccel
erat
ion
(ms
2 )
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 26 e dynamic responses of the sprung mass accelerationof the hybrid suspension in semiactive mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 27e power spectrums of the sprungmass acceleration ofthe hybrid suspension in semiactive mode
Table 7 e sprung mass acceleration RMS of the hybrid sus-pension in semiactive mode
Indicators Passivesuspension
Semiactivesuspension
Hybridsuspension
aw (ms2) 13378 11536 11182
14 Shock and Vibration
6 Conclusion
(1) A ball screw actuator andMR damper are introducedinto vehicle suspension system and a new kind ofhybrid active suspension structure is put forwarde ball screw actuator is prototyped and thefunctional relationship between the back-EMF co-ecient the electromagnetic torque coecient of themotor and the suspension vibration speed is ob-tained by test analyses And the active output me-chanical properties of the ball screw actuator aretested and the results show that the actuator hasgood active output force characteristics
(2) e inshyuences of the variable damping value of thesuspension system on the riding comfort handlingand stability and energy consumption characteristicsof the hybrid suspension in the active control modeare analyzeden the optimal damping values of thehybrid suspension at dierent vehicle speeds aredesigned e eects of electromagnetic dampingforce on the actual semiactive force and the systemenergy of the suspension in the semiactive controlmode are analyzed and then the hybrid suspensionwith semiactive mode which has electromagneticdamping force feedback adjustment is designed Onthis basis a multimode coordinated control strategyfor the hybrid suspension is designed
(3) e damping performance and energy consumptioncharacteristics of the hybrid suspension under cyclicdriving condition are simulated by MATLABSimulink software and the results show that whenin the urban circulation unit compared with activesuspension aw and DTLrms of the hybrid suspensionare reduced by 1322 and 840 respectively Andwhen in the suburban circulation unit comparedwith the active suspension aw and DTLrms of thehybrid suspension are reduced by 1199 and 732respectively e pure energy of the active suspen-sion system during the entire operation cycle is
minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid sus-pension system basically realizes energy self-poweredin theory
(4) e eectiveness verication test of the hybrid sus-pension in active mode and semiactive mode controlis carried out and the results show that when thehybrid suspension is actively controlled comparedwith active suspension aw of the hybrid suspension inactive mode is reduced by 1432 When the hybridsuspension is semiactively controlled compared withsemiactive suspension aw of the hybrid suspension inactive mode is reduced by 307 e test and sim-ulation results are basically consistent and the testveries the correctness of the simulation
Data Availability
e data used to support the ndings of this study are in-cluded within the supplementary information les And thedata used to support the ndings of this study are availablefrom the corresponding author upon request
Conflicts of Interest
e authors declare that there are no conshyicts of interestregarding the publication of this paper
Acknowledgments
is work was supported by the National Natural ScienceFoundation of China (Grant no 51775426) Service LocalSpecial Program Support Project of Shaanxi ProvincialEducation Department (Grant no 17JF017) and XirsquoanScience and Technology Program Funding Project (Grantno 2017079CGRC042-XAKD007)
Supplementary Materials
e supplementary materials are this articlersquos experimentaldata including (1) MR damperrsquos characteristic test data (2)the motor counter electromotive forcersquos peak test data (3)the ball screw actuatorrsquos active output force test data (4) thehybrid suspensionrsquos sprung mass acceleration test data inactive mode (5) the hybrid suspensionrsquos sprung mass ac-celeration test data in semiactive mode and (6) the hybridsuspensionrsquos energy consumption power and energy re-generation power test data (Supplementary Materials)
References
[1] M Montazeri-Gh and O Kavianipour ldquoInvestigation of theactive electromagnetic suspension system considering hybridcontrol strategyrdquo Proceedings of the Institution of MechanicalEngineers Part C Journal of Mechanical Engineering Sciencevol 228 no 10 pp 1658ndash1669 2013
[2] H E Tseng and D Hrovat ldquoState of the art survey active andsemi-active suspension controlrdquo Vehicle System Dynamicsvol 53 no 7 pp 1034ndash1062 2015
0 1 2 3 4 50
20
40
60
80
Time (s)
Pow
er (W
)
Energy consumption power in active modeEnergy regenerative power in semiactive modeEnergy consumption power in semiactive mode
Figure 28 e energy consumption power and energy-re-generative power of the hybrid suspension
Shock and Vibration 15
[3] F Kou J Du Z Wang D Li and J Xu ldquoNonlinear modelingand coordinate optimization of a semi-active energy re-generative suspension with an electro-hydraulic actuatorrdquoChina Mechanical Engineering vol 28 no 14 pp 1701ndash17072017
[4] S H Zareh A Sarrafan A A A Khayyat and A ZabihollahldquoIntelligent semi-active vibration control of eleven degrees offreedom suspension system using magnetorheologicaldampersrdquo Journal of Mechanical Science and Technologyvol 26 no 3 pp 323ndash334 2012
[5] S A Chen X Li L J Zhao Y X Wang and Y B KimldquoDevelopment of a control method for an electromagneticsemi-active suspension reclaiming energy with varying chargevoltage in stepsrdquo International Journal of Automotive Tech-nology vol 16 no 5 pp 765ndash773 2015
[6] K Nakano Y Suda and S Nakadai ldquoSelf-powered activevibration control using a single electric actuatorrdquo Journal ofSound and Vibration vol 260 no 2 pp 213ndash235 2003
[7] K Nakano ldquoCombined type self-powered active vibrationcontrol of truck cabinsrdquo Vehicle System Dynamics vol 41no 6 pp 449ndash473 2004
[8] K Huang Y C Zhang F Yu and Y H Gu ldquoCoordinateoptimization for synthetical performance of electrical energy-regenerative active suspensionrdquo Journal of Shanghai JiaotongUniversity vol 43 no 2 pp 226ndash230 2009
[9] K Huang F Yu and Y C Zhang ldquoActive control of energy-regenerative electromagnetic suspension based on energy flowanalysisrdquo Journal of Shanghai Jiaotong University vol 45no 67 pp 1068ndash1073 2011
[10] D S Huang J Q Zhang Y L Liu L Yi and X Y WangldquoPerformance of a novel energy-regenerative active suspen-sion systemrdquo Journal of Chongqing University (English Edi-tion) vol 14 no 3 pp 109ndash118 2015
[11] B Ebrahimi H Bolandhemmat M B Khamesee andF Golnaraghi ldquoA hybrid electromagnetic shock absorber foractive vehicle suspension systemsrdquo Vehicle System Dynamicsvol 49 no 1-2 pp 311ndash332 2011
[12] S Tang L Chen R Wang X Sun and D Shi ldquoResearch onoptimal control of active suspension based on damping multi-modal switchingrdquo Journal of Guangxi University (NaturalScience) vol 39 no 2 pp 300ndash307 2014
[13] R Wang X Ma R Ding X Meng and L Chen ldquoResearchof multi-mode switching control system for hybrid sus-pension based on model referencerdquo Transactions of theChinese Society for Agricultural Machinery vol 48 no 7pp 353ndash360 2017
[14] R Wang Y Qian R Ding X Meng and J Xie ldquoDesign andtests for damping-stiffness of a hybrid electromagnetic sus-pension based on LQGrdquo Journal of Vibration and Shockvol 37 no 3 pp 61ndash65 2017
[15] B Vanavil K K Chaitanya and A S Rao ldquoImproved PIDcontroller design for unstable time delay processes based ondirect synthesis method and maximum sensitivityrdquo Taylorand Francis vol 46 no 8 pp 1349ndash1366 2015
[16] I Mihai and F Andronic ldquoBehavior of a semi-active sus-pension system versus a passive suspension system on anuneven road surfacerdquo Mechanics vol 20 no 1 pp 64ndash692014
[17] D Ngoduy ldquoLinear stability of a generalized multi-anticipative car following model with time delaysrdquo Com-munications in Nonlinear Science and Numerical Simulationvol 22 no 1 pp 420ndash426 2015
[18] H Zhang E Wang F Min R Subash and C Su ldquoSkyhook-based semi-active control of full-vehicle suspension with
magneto-rheological dampersrdquo Chinese Journal of Mechan-ical Engineering vol 26 no 3 pp 498ndash505 2013
[19] F Kou ldquoAn experimental study on the dynamic character-istics of vehicle semi-active seat suspension with magneto-rheological damperrdquo Automotive Engineering vol 37 no 11pp 1346ndash1352 2015
[20] Z Feng S Chen and Y Liang ldquoAn experimental study on thedynamic characteristics of a megneto-rheological semi-activesuspensionrdquoAutomotive Engineering vol 35 no 1 pp 72ndash772013
[21] Z Li and L Xu A New Type of Magnetorheological Damperand Semi-Active coNtrol Design-eory Science Press BeijingChina 2012
[22] B L J Gysen J J H Paulides J L G Janssen andE A Lomonova ldquoActive electromagnetic suspension systemfor improved vehicle dynamicsrdquo IEEE Transactions on Ve-hicular Technology vol 59 no 3 pp 1156ndash1163 2010
[23] Q N Wang S S Liu W H Wang and H Wei ldquoStructuredesign and parameter matching of ball-screw regenerativedamperrdquo Journal of Jilin University (Engineering and Tech-nology Edition) vol 42 no 5 pp 1100ndash1106 2012
[24] S Chen R He and S Lu ldquoEvaluating system of reclaimingenergy suspension comprehensive performancerdquo Trans-actions of the Chinese Society for Agricultural Machineryvol 37 no 7 pp 14ndash18 2006
[25] Y Fan Control Research on Vehicle Suspension with Electro-Hydrostatic Actuator Xirsquoan University of Science and Tech-nology Xirsquoan China 2017
[26] Z Yu Automobile -eory China Machine Press BeijingChina 2009
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Navigation and Observation
International Journal of
Hindawi
wwwhindawicom Volume 2018
Advances in
Multimedia
Submit your manuscripts atwwwhindawicom
e vibration velocity of the hybrid suspension and themotor rotating speed meet the following relationships
n ΔvLmiddot 60 (14)
Δv _x2 minus _x1 (15)
where Δv is the suspension vibration velocity and L is ballscrew lead
At dierent suspension vibration velocity the motorback-EMF coecient can be obtained from formulae(13)sim(15) and the tting curve of the motor back-EMFcoecient is shown in Figure 7
In Figure 7 the tting relationship between ke and Δv isexpressed as
ke 00299Δv3 minus 00542Δv2 + 00253Δv + 00049 (16)
en kT and ke meet the following relationships
kT 30πke (17)
e relationship between kT and Δv can be obtainedfrom formulae (16) and (17)
kT 02855Δv3 minus 05176Δv2 + 02416Δv + 00468 (18)
When testing the active output force of the ball screwactuator the ball screw actuator is powered by the same 72Vconstant voltage source as the battery pack terminal voltageBy adjusting PWM duty ratio the input voltage and currentof the motor are controlled by the controller and as a resultthe output force of the motor is gained e relationshipbetween the active output force of the ball screw actuatorand duty ratio is shown in Figure 8
From Figure 8 it can be seen that the ball screw actuatorhas good active output force characteristics and the non-linear relationship between the output force and the dutyratio is consistent with the nonlinear characteristics of themotor
332 Ball Screw Actuator Mathematical Model When themotor inductance is ignored in the active control of thehybrid suspension input voltage E and the output torqueTmof the motor meet the following relationships
E u + Izr (19)
Tm kT middot Iz (20)
u 60 _x2 minus _x1( )
Lmiddot ke (21)
Tm F middot L2π
(22)
ke kT middot π30
(23)
400 800 1200 1600 2000 2400 2800 3200 3600 40000
10
20
30
40
50
60
70
Rotating speed (rmin)
The p
eak
valu
e of c
ount
erel
ectr
omot
ive f
orce
(V)
The peak of counter electromotive forceTest fitting curve
Figure 6 e peak value curve of counter electromotive force ofthe motor
008 020 032 044 056 068 08000055
00060
00065
00070
00075
00080
00085
Suspension vibration speed (ms)
Back
-EM
F co
effic
ient
(Vr
middotmin
ndash1)
Test value data pointsTest fitting curve
Figure 7 e motor back-EMF coecient curve
ndash100 ndash50 0 50 100ndash600
ndash400
ndash200
0
200
400
600
Duty cycle ()
Act
uato
r act
ive o
utpu
t for
ce (N
)
Actuator active output forceTest fitting curve
Figure 8 Testing curve of the actuator active output force
Shock and Vibration 5
where u is induced electromotive force Iz is motor currentand r is internal resistance of the motor
In the active control of the hybrid suspension the in-stantaneous energy consumption power of the ball screwactuator can be expressed as
Pz E middot Iz
ηz (24)
0e instantaneous energy consumption power andconsumption energy can be obtained from formulae(19)sim(24) as follows
Pz _x2 minus _x1 1113857 middot F + FL2πkT 1113857
2middot r1113960 1113961
ηz (25)
Wz 1113946t
0Pz dt (26)
where Pz is motor instantaneous energy consumptionpower ηz is transfer efficiency of ball screw actuator andWzis motor energy consumption
In the energy regeneration of the hybrid suspension theelectromagnetic damping force generated by the ball screwactuator is expressed as
Fs minus2πkT
L1113888 1113889
2
middot_x2 minus _x1 1113857
R + rηz (27)
where Fs is the electromagnetic damping force and R isexternal resistance of the motor
0e instantaneous energy-regenerative power and regen-erative energy of the ball screw actuator can be expressed as
Pk 2πke
L1113888 1113889
2
middot_x2 minus _x1 1113857
2
R + rηz (28)
Wk 1113946t
0Pk dt (29)
where Pk is instantaneous energy-regenerative power andWk is regenerative energy
4 Multimode Coordination Control of HybridActive Suspension
41 -e Active Mode of Damping Switching Control 0esprung mass acceleration is the main evaluation index ofvehicle riding comfort and the dynamic tire load is closelyrelated to vehicle handling and stability An active controlmodel of the hybrid suspension is established by MATLABSimulink software to simulate and analyze the influence ofthe variable damping for the hybrid active suspension onvehicle riding comfort and handling and stability at differentvehicle speeds During the simulation the range of vehiclespeed v is 0ndash120 kmh and the vehicle speed is taken every10 kmh 0e variable damping range of the suspension is200ndash2000 Nmiddotsm and the value of the variable damping istaken every interval 100 Nmiddotsm 0e simulation time is 10 sand the value of r is 05 Ω 0e value of ηz is 097 and thevalue of ηb is 098 0e value of csky is 2000 Nmiddotsm
When the vehicle speed is 30 kmh and 100 kmh re-spectively the RMS of the sprung mass acceleration of thevehicle (aw) and the RMS of the dynamic tire load (DTLrms)change with the variable damping of the hybrid activesuspension as shown in Figures 9 and 10
From Figures 9 and 10 it can be seen that when thevehicle speed is 30 kmh the variable damping values thatmake aw of vehicle and DTLrms minimum are 400 Nmiddotsm and1000 Nmiddotsm respectively and when the vehicle speed is100 kmh the variable damping values that make aw ofvehicle and DTLrms minimum are 500 Nmiddotsm and 1100 Nmiddotsm respectively 0erefore at a certain vehicle speed thevariable damping of the hybrid active suspension cannotmake the best of the vehicle riding comfort and handling andstability at the same time
Suspension performance indexes include the sprung massacceleration suspension working space and dynamic tireload In this paper in order to balance vehicle riding comfortand handling and stability when choosing the variabledamping values of the hybrid suspension in the active controlfor the sprung mass acceleration suspension working spaceand dynamic tire load of the hybrid active suspension thequantitative normalizations and comparative analyses aredone 0at is at the same vehicle speed compared with thepassive suspension the improvement amplitudes of eachperformance index of the hybrid active suspension aremultiplied by different quantification factors and summedAnd the larger the sum the better the dynamic performanceof vehicle Among them the quantification factors of aw theRMS value of suspension working space (SWSrms) andDTLrms are 1 02703 and 01443 respectively [24 25] Andwhen the vehicle speed is 30 kmh and 100 kmh respectivelythe dynamic performance and the active control energyconsumption of the hybrid active suspension change with thevariable damping as shown in Figures 11 and 12
From Figures 11 and 12 when the vehicle speed is 30 kmh the variable damping value of the hybrid active suspensionis 800 Nmiddotsm which makes the vehicle dynamic performancethe best and the active energy consumption the least Whenthe vehicle speed is 100 kmh the variable damping value ofthe hybrid active suspension is 1000 Nmiddotsm which makes thevehicle dynamic performance the best and active energyconsumption the least 0erefore when the vehicle speed is30 kmh and 100 kmh respectively the optimal dampingvalues of the hybrid active suspension are 800 Nmiddotsm and1000 Nmiddotsm respectively When the vehicle speed is 0ndash120 kmh the optimal damping values of the hybrid activesuspension at different vehicle speeds are shown in Figure 13If the damping value which makes the vehicle dynamicperformance the best is different from the damping valuewhich makes the active control energy consumption the leastthe damping value which makes the vehicle dynamic per-formance the best is selected as the optimal damping value ofthe hybrid active suspension at the vehicle speed
When vehicle is in an accelerating or decelerating stateits speed changes rapidly and the range of change is wide sothe vehicle speed value is not easily detected in real time andin order to reduce the energy consumption of the hybridsuspension active controlled and to improve the vehicle
6 Shock and Vibration
riding comfort and handling and stability a variabledamping switching control strategy of the hybrid activesuspension is designed as follows
Fz minus csky middot _x2cs c0 + ck
_v 0
Fz minus csky middot _x2cs c0
_vne 0
(30)
where _v is vehicle acceleration and csky is sky-hook coecient
42 e Semiactive Mode of Feedback Adjustment of Elec-tromagnetic Damping Force MR damper can eectivelyperform semiactive control at ( _x2 minus _x1) _x2 gt 0 so the idealsemiactive control state for hybrid suspension is
Fb minus csky middot _x2 _x2 minus _x1( ) _x2 gt 0
0 _x2 minus _x1( ) _x2 le 0
(31)
From equation (31) the ideal semiactive control force ofthe hybrid suspension minus csky middot _x2 is only related to _x2 when
Vehicle dynamics performanceEnergy consumption of hybrid suspension
Figure 12 Relationship between the variable damping and vehicledynamic performance at 100 kmh speed
20 40 60 80 100 120750
800
850
900
950
1000
1050
Vehicle speed (kmh)
Shoc
k ab
sorb
er d
ampi
ngco
effic
ient
(Nmiddots
m)
Figure 13 Suspension optimal damping at dierent speeds
Shock and Vibration 7
the sky-hook coecient csky is constant However at thistime the ball screw actuator as a power feeding devicegenerates the electromagnetic damping force Fs and acts onthe suspension so that the actual semiactive control force ofthe suspension is dierent from the ideal semiactive controlforce minus csky middot _x2 In this paper the semiactive control modelof the hybrid suspension is established and the changeeects of the dierent output forces on the vehicle ridingcomfort and handling stability are analyzed by MATLABSimulink software e simulation speed is 70 kmh thesimulation time is 5 s and the value of R is 075 Ω edamping comparison of the hybrid suspension in semiactivecontrol is shown in Figure 14
Figure 14 shows that compared with the ideal semi-active control force the actual semiactive control force ofthe hybrid suspension shyuctuates violently and the abso-lute value of the actual semiactive control force is greaterthan the absolute value of the ideal semiactive control force|csky middot _x2| at certain times And a drastic change in theactual semiactive control force makes the suspension notreach ideal semiactive control eect Using electromag-netic damping force feedback adjustment to reduce thedierence between the ideal semiactive control forceminus csky middot _x2 and the actual semiactive control force themethod is as follows
When |csky middot _x2|gt |Fs| the semiactive control force of thehybrid suspension is provided by both the MR damper andthe ball screw actuator and at this point the controllerinputs a controllable current Ik to the MR damper so thatthe Fk output by the MR damper is minus csky middot _x2 minusFs Andwhen |csky middot _x2|le |Fs| the semiactive control force of thehybrid suspension is the Fs which is output by the ball screwactuator and at this point there is no controllable current Ikinput to theMR damper and the function of theMR damperis equivalent to a traditional shock absorber erefore thesemiactive control of the hybrid suspension does not havethe dead zone of traditional electromagnetic semiactivesuspension which helps to improve the semiactive controleect of the hybrid suspension
When there is feedback adjustment the semiactivecontrol force of the hybrid suspension is
Fb Fs Fs
∣∣∣∣∣∣∣∣ge csky middot _x2∣∣∣∣∣
∣∣∣∣∣
minus csky middot _x2 Fs∣∣∣∣∣∣∣∣lt csky middot _x2∣∣∣∣∣
∣∣∣∣∣
(32)
When there is feedback adjustment the Fk output by theMR damper is
Fk 0 Fs
∣∣∣∣∣∣∣∣ge csky middot _x2∣∣∣∣∣
∣∣∣∣∣
minus csky middot _x2 minusFs Fs∣∣∣∣∣∣∣∣lt csky middot _x2∣∣∣∣∣
∣∣∣∣∣
(33)
From equations (31)sim(33) when the electromagneticdamping force feedback adjustment is used the |Fk| outputby theMR damper decreases and when |Fk| decreases it canbe known from equations (8) and (9) that the energyconsumption of the MR damper decreases with it
e comparison of the semiactive control force of thehybrid suspension with or without the electromagneticdamping force feedback adjustment is shown in Figure 15
From Figure 15 the RMS of the ideal semiactive controlforce of the hybrid suspension is 3276 N and when there isno electromagnetic damping force feedback adjustmentthe RMS of the actual semiactive control force of thesuspension is 4041 N and the dierence between the actualsemiactive force of the suspension and the ideal semiactiveforce is 2335 When there is electromagnetic dampingforce feedback adjustment the RMS of the actual semi-active control force of the suspension is 3593 N and thedierence between the actual semiactive force of the sus-pension and the ideal semiactive force is 968 ereforewhen there is electromagnetic damping force feedbackadjustment the actual semiactive control force of thesuspension has a smaller shyuctuation amplitude whichhelps to improve the semiactive control eect of the hybridsuspension
e dynamic responses of the hybrid suspension with orwithout electromagnetic damping force feedback adjust-ment are shown in Figure 16 Among them the damper ofthe passive suspension is the original damper of the vehicleand its damping value is 1600Nmiddotsm
Table 2 shows the response RMS values of the hybridsuspension in semiactive control
From Table 2 compared with the passive suspensionwhen there is electromagnetic damping force feedbackadjustment aw SWSrms and DTLrms of the hybrid sus-pension are reduced by 1698 432 and 1068 re-spectively and compared with the nonfeedback semiactivecontrol when the feedback semiactive control is performedaw SWSrms and DTLrms of the hybrid suspension are re-duced by 252 863 and 671 respectively
From equations (9) and (29) the total system energy ofthe hybrid suspension in semiactive control is
W1 Wk minusWb (34)
whereW1 is the total system energy of the hybrid suspensionin semiactive control
0 1 2 3 4 5ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Hyb
rid su
spen
sion
forc
e (N
)
072 076 080
300
600
Ideal semiactive forceActual semiactive forceElectromagnetic damping force
Figure 14 Damping comparison of the hybrid suspension insemiactive control
8 Shock and Vibration
0 1 2 3 4 5ndash10
ndash5
0
5
10
15
Time (s)
Spru
ng m
ass a
ccel
erat
ion
(ms
2 )
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(a)
0 1 2 3 4 5ndash004
ndash002
000
002
004
006
008
Time (s)
Susp
ensio
n w
orki
ng sp
ace (
m)
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(b)
0 1 2 3 4 5ndash3000
ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Dyn
amic
tire
load
(N)
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(c)
Figure 16 e dynamic responses of the hybrid suspension in semiactive control (a) e response curves of sprung mass acceleration (b)e response curves of suspension working space (c) e response curves of dynamic tire load
0 1 2 3 4 5ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Hyb
rid su
spen
sion
forc
e (N
)
072 076 080
300
600
Ideal semiactive forceWithout feedback semiactive forceWith feedback semiactive force
Figure 15 Semiactive force of the hybrid suspension
Shock and Vibration 9
From equation (33) when the hybrid suspension sem-iactive controlled the curves of the total system energychange over time are shown in Figure 17
From Figure 17 when there is the nonfeedback semiactivecontrol the total system energy of the hybrid suspension is60 J And when there is the feedback semiactive control thetotal system energy of the hybrid suspension is 307 J
43 -e Design of Multimode Coordination ControllerWhen the hybrid suspension is actively controlled thesuspension has good vibration isolation performance buthigh energy consumption And when the hybrid suspensionis semiactively controlled the suspension has good eco-nomic performance but the control has limitations Con-sidering that the ball screw actuator can realize active controlof the hybrid suspension in any suspension state a multi-mode coordinated control strategy of the hybrid suspensionis designed
minus csky middot _x2 Fs1113868111386811138681113868
1113868111386811138681113868gt csky middot _x2
11138681113868111386811138681113868
11138681113868111386811138681113868
⎧⎪⎨
⎪⎩
cs c0
⎧⎪⎪⎪⎨
⎪⎪⎪⎩
_x2 minus _x1 1113857 _x2 gt 0
F Fz minus csky middot _x2
cs c0 + ck1113896 _x2 minus _x1 1113857 _x2 le 0
⎧⎪⎪⎪⎪⎪⎪⎪⎪⎪⎨
⎪⎪⎪⎪⎪⎪⎪⎪⎪⎩
(35)
0e frame diagram of the multimode coordinated controlstrategy of the hybrid suspension is shown in Figure 18
From equations (9) (12) (26) and (29) the total systemenergy of the hybrid suspension in multimode coordinatedcontrol is
W2 Wk minusWb minusWZ minusWc (36)
whereW2 is the total system energy of the hybrid suspensionin multimode coordinated control
A hybrid active suspension simulation model is estab-lished by using MATLABSimulink software From Fig-ure 11 when the hybrid suspension is actively controlled theenergy consumption of the MR damper is mainly affected bythe vehicle speed In order to verify the vibration isolationperformance and energy consumption performance of thehybrid active suspension in the cyclic driving conditionsthis article simulates vehicle urban and suburban conditionswhich is based on GBT 19233-2003 ldquoLight Vehicle FuelConsumption Test Methodrdquo among them the urban con-ditions include four cycle units each cycle time is 195 s thesuburban conditions include one cycle unit and the cycletime is 400 s [26] And the schematic diagram of the urbanconditions unit is shown in Figure 19
In order to simulate the vehicle acceleration signal avehicle speed variation model in different cycle units isestablished by using the signal builder function module inSimulink software On this basis the vehicle accelerationmodel in different cycle units is obtained and the vehicleacceleration model can be used as the switching controlmodel for the variable damping when the hybrid suspensionis actively controlled 0e shock absorbers of active sus-pension and passive suspension adopt the original damper ofvehicle and the damping value c1 is 1600 Nmiddotsm Howeverbecause of the long simulation time the dynamic responsecurves of the hybrid active suspension in urban and sub-urban conditions cannot be displayed Figure 20 shows thedynamic response curve of the hybrid active suspension inthe 145ndash175 s in the urban circulation unit among themvehicle in the 145ndash155 s is in a constant speed vehicle in the155ndash163 s is in the deceleration state and vehicle in the163ndash175 s is in other constant speed
0e dynamic response RMS values of the hybrid sus-pension in urban and suburban circulation units are shownin Tables 3 and 4 respectively
From Figure 20 and Tables 3 and 4 the dynamic re-sponses are good when the hybrid suspension is multimodecoordinated control When in the urban circulation unitcompared with the passive suspension aw SWSrms andDTLrms of the hybrid suspension are reduced by 3943225 and 2081 respectively Compared with the activesuspension aw and DTLrms of the hybrid suspension arereduced by 1322 and 840 respectively but SWSrms isincreased by 805 When in the suburban circulation unitcompared with the passive suspension aw SWSrms andDTLrms of the hybrid suspension are reduced by 3916314 and 1955 respectively Compared with the activesuspension aw and DTLrms of the hybrid suspension arereduced by 1199 and 732 respectively but SWSrms isincreased by 845
0e system energy of the hybrid suspension in urban andsuburban circulation units are shown in Figures 21 and 22respectively
Table 2 0e response root mean square values of the hybridsuspension in semiactive control
Without feedback semiactive controlWith feedback semiactive control
Figure 17 0e system energy of the hybrid suspension in semi-active control
10 Shock and Vibration
en the system energy values of the hybrid suspensionin urban and suburban circulation units are shown inTable 5
From Table 5 when in the urban circulation unit thesystem energy of the active suspension and hybrid sus-pension is minus2165 J and 38 J respectively And when in thesuburban circulation unit the system energy of the activesuspension and hybrid suspension is minus15071 J and minus122 Jrespectively e entire operation cycle includes 4 urbancycle units and 1 suburban cycle unit so the pure energy ofthe active suspension system during the entire operationcycle is minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid suspensionsystem basically realizes energy self-powered in theory
5 Test and Analysis
In order to verify the damping eect of the hybrid sus-pension system in active mode and semiactive mode ahybrid suspension vibration test system is designed as
Suspension controllerSpeed of sprung mass x2
Speed of unsprung mass x1Vehicle acceleration v
Vehicle suspension
Ball screw actuator
Output Fk
Active control force Fz
Ball screw actuator
Output Iz
Output Ik
Electromagnetic damping force Fs
Ideal semiactive control force FL
Output semiactive controlforce Fb = Fs
Output semiactive controlforce Fb = FL = Fk + Fs
x2 ndash x1
(x2 ndash x1) x2 gt 0
(x2 ndash x1) x2 le 0
Output Ik
MR damper
Adjust the duty cycle in real timeaccording to the relationshipbetween active output force
and duty cycle
No controllablecurrent output
MR damper
v ne 0
According to vand ∆v
v = 0
MR damper
Adjustable damping cs = c0 + ck
ndashcsky middot x2
ndashcsky middot x2
|Fs| ge |FL|
|Fs| lt |FL|
Figure 18 e frame diagram of the multimode coordinated control strategy of the hybrid suspension
Vehicle shifting
00
15
10
30
30
20
45
40
50
60
60
75 90 105 120Time (s)
Veh
icle
spee
d (k
mh
)
135 150 165 180 195
Figure 19 e schematic diagram of the urban conditions unit
Shock and Vibration 11
shown in Figure 23 During the test the MR damper has nocontrollable current input in the active control mode of thehybrid suspension and its damping value is always 800Nmiddotsm
e passive suspension semiactive suspension and activesuspension all adopt the original damper of the vehicle andthe damping value c1 is 1600Nmiddotsm
Figure 20e dynamic responses of the hybrid suspension (a)e response curves of sprungmass acceleration (b)e response curves ofsuspension working space (c) e response curves of dynamic tire load
12 Shock and Vibration
Because of the limitation of test conditions only thesprung mass acceleration dynamic response of the hybridsuspension is measured in this test e dynamic response ofthe sprung mass acceleration of the hybrid suspension inactive mode under random road is shown in Figure 24 epower spectrum of the sprung mass acceleration of thehybrid suspension in active mode is shown in Figure 25
Table 6 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in active mode
From Table 6 compared with passive suspension aw ofthe hybrid suspension in active mode is reduced by 3945and compared with active suspension aw of the hybridsuspension in active mode is reduced by 1432 And fromFigure 25 when the hybrid suspension is actively controlledcompared with passive suspension the vibration isolationperformance of the suspension in the low frequency andlow-frequency resonance areas is similar to that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withactive suspension the vibration isolation performance of thesuspension in the low frequency high frequency and res-onance regions is slightly better
e dynamic response of the sprung mass acceleration ofthe hybrid suspension in semiactive mode under randomroad is shown in Figure 26 e power spectrum of thesprung mass acceleration of the hybrid suspension insemiactive mode is shown in Figure 27
Table 7 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in semiactive mode
From Table 7 compared with passive suspension aw ofthe hybrid suspension in semiactive mode is reduced by1642 Compared with semiactive suspension aw of thehybrid suspension in active mode is reduced by 307 Andfrom Figure 27 when the hybrid suspension is semiactivelycontrolled compared with passive suspension the vibration
Table 3 e dynamic response root mean square values of thehybrid suspension in urban circulation unit
Figure 21 e system energy of the hybrid suspension in urbancirculation units
0 100 200 300 400ndash20000
ndash15000
ndash10000
ndash5000
0
5000
Time (s)
Susp
ensio
n sy
stem
ener
gy (J
)
Active suspensionHybrid suspension
Figure 22e system energy of the hybrid suspension in suburbancirculation units
Table 5 e system energy values of the hybrid suspension (J)
Cycle unit Active suspension Hybrid suspensionUrban minus2165 38Suburban minus15071 minus122
Figure 23 e test system of the hybrid active suspension
Shock and Vibration 13
isolation performance of the suspension in the low frequencyand low-frequency resonance areas is worse than that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withsemiactive suspension the vibration isolation performance ofthe suspension in the low frequency high frequency andresonance regions is slightly better
e energy consumption power and energy-regenerativepower of the hybrid suspension are shown in Figure 28
From Figure 28 in the active mode the average powerconsumption of the ball screw actuator is 2561W In thesemiactive mode the average regenerative power of the ballscrew actuator is 2696W and the average of theMR damperenergy consumption power is 178W erefore the pureaverage power consumption of the hybrid suspension isminus043W and the test results are basically consistent with thesimulation
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spr
ung
mas
s acc
eler
atio
n (m
s2 )
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 24 e dynamic response of the sprung mass accelerationof the hybrid suspension in active mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 25 e power spectrum of the sprung mass acceleration ofthe hybrid suspension in active mode
Table 6 e sprung mass acceleration RMS of the hybrid sus-pension in active mode
Indicators Passivesuspension
Activesuspension
Hybridsuspension
aw (ms2) 13378 09423 08074
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spru
ng m
ass a
ccel
erat
ion
(ms
2 )
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 26 e dynamic responses of the sprung mass accelerationof the hybrid suspension in semiactive mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 27e power spectrums of the sprungmass acceleration ofthe hybrid suspension in semiactive mode
Table 7 e sprung mass acceleration RMS of the hybrid sus-pension in semiactive mode
Indicators Passivesuspension
Semiactivesuspension
Hybridsuspension
aw (ms2) 13378 11536 11182
14 Shock and Vibration
6 Conclusion
(1) A ball screw actuator andMR damper are introducedinto vehicle suspension system and a new kind ofhybrid active suspension structure is put forwarde ball screw actuator is prototyped and thefunctional relationship between the back-EMF co-ecient the electromagnetic torque coecient of themotor and the suspension vibration speed is ob-tained by test analyses And the active output me-chanical properties of the ball screw actuator aretested and the results show that the actuator hasgood active output force characteristics
(2) e inshyuences of the variable damping value of thesuspension system on the riding comfort handlingand stability and energy consumption characteristicsof the hybrid suspension in the active control modeare analyzeden the optimal damping values of thehybrid suspension at dierent vehicle speeds aredesigned e eects of electromagnetic dampingforce on the actual semiactive force and the systemenergy of the suspension in the semiactive controlmode are analyzed and then the hybrid suspensionwith semiactive mode which has electromagneticdamping force feedback adjustment is designed Onthis basis a multimode coordinated control strategyfor the hybrid suspension is designed
(3) e damping performance and energy consumptioncharacteristics of the hybrid suspension under cyclicdriving condition are simulated by MATLABSimulink software and the results show that whenin the urban circulation unit compared with activesuspension aw and DTLrms of the hybrid suspensionare reduced by 1322 and 840 respectively Andwhen in the suburban circulation unit comparedwith the active suspension aw and DTLrms of thehybrid suspension are reduced by 1199 and 732respectively e pure energy of the active suspen-sion system during the entire operation cycle is
minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid sus-pension system basically realizes energy self-poweredin theory
(4) e eectiveness verication test of the hybrid sus-pension in active mode and semiactive mode controlis carried out and the results show that when thehybrid suspension is actively controlled comparedwith active suspension aw of the hybrid suspension inactive mode is reduced by 1432 When the hybridsuspension is semiactively controlled compared withsemiactive suspension aw of the hybrid suspension inactive mode is reduced by 307 e test and sim-ulation results are basically consistent and the testveries the correctness of the simulation
Data Availability
e data used to support the ndings of this study are in-cluded within the supplementary information les And thedata used to support the ndings of this study are availablefrom the corresponding author upon request
Conflicts of Interest
e authors declare that there are no conshyicts of interestregarding the publication of this paper
Acknowledgments
is work was supported by the National Natural ScienceFoundation of China (Grant no 51775426) Service LocalSpecial Program Support Project of Shaanxi ProvincialEducation Department (Grant no 17JF017) and XirsquoanScience and Technology Program Funding Project (Grantno 2017079CGRC042-XAKD007)
Supplementary Materials
e supplementary materials are this articlersquos experimentaldata including (1) MR damperrsquos characteristic test data (2)the motor counter electromotive forcersquos peak test data (3)the ball screw actuatorrsquos active output force test data (4) thehybrid suspensionrsquos sprung mass acceleration test data inactive mode (5) the hybrid suspensionrsquos sprung mass ac-celeration test data in semiactive mode and (6) the hybridsuspensionrsquos energy consumption power and energy re-generation power test data (Supplementary Materials)
References
[1] M Montazeri-Gh and O Kavianipour ldquoInvestigation of theactive electromagnetic suspension system considering hybridcontrol strategyrdquo Proceedings of the Institution of MechanicalEngineers Part C Journal of Mechanical Engineering Sciencevol 228 no 10 pp 1658ndash1669 2013
[2] H E Tseng and D Hrovat ldquoState of the art survey active andsemi-active suspension controlrdquo Vehicle System Dynamicsvol 53 no 7 pp 1034ndash1062 2015
0 1 2 3 4 50
20
40
60
80
Time (s)
Pow
er (W
)
Energy consumption power in active modeEnergy regenerative power in semiactive modeEnergy consumption power in semiactive mode
Figure 28 e energy consumption power and energy-re-generative power of the hybrid suspension
Shock and Vibration 15
[3] F Kou J Du Z Wang D Li and J Xu ldquoNonlinear modelingand coordinate optimization of a semi-active energy re-generative suspension with an electro-hydraulic actuatorrdquoChina Mechanical Engineering vol 28 no 14 pp 1701ndash17072017
[4] S H Zareh A Sarrafan A A A Khayyat and A ZabihollahldquoIntelligent semi-active vibration control of eleven degrees offreedom suspension system using magnetorheologicaldampersrdquo Journal of Mechanical Science and Technologyvol 26 no 3 pp 323ndash334 2012
[5] S A Chen X Li L J Zhao Y X Wang and Y B KimldquoDevelopment of a control method for an electromagneticsemi-active suspension reclaiming energy with varying chargevoltage in stepsrdquo International Journal of Automotive Tech-nology vol 16 no 5 pp 765ndash773 2015
[6] K Nakano Y Suda and S Nakadai ldquoSelf-powered activevibration control using a single electric actuatorrdquo Journal ofSound and Vibration vol 260 no 2 pp 213ndash235 2003
[7] K Nakano ldquoCombined type self-powered active vibrationcontrol of truck cabinsrdquo Vehicle System Dynamics vol 41no 6 pp 449ndash473 2004
[8] K Huang Y C Zhang F Yu and Y H Gu ldquoCoordinateoptimization for synthetical performance of electrical energy-regenerative active suspensionrdquo Journal of Shanghai JiaotongUniversity vol 43 no 2 pp 226ndash230 2009
[9] K Huang F Yu and Y C Zhang ldquoActive control of energy-regenerative electromagnetic suspension based on energy flowanalysisrdquo Journal of Shanghai Jiaotong University vol 45no 67 pp 1068ndash1073 2011
[10] D S Huang J Q Zhang Y L Liu L Yi and X Y WangldquoPerformance of a novel energy-regenerative active suspen-sion systemrdquo Journal of Chongqing University (English Edi-tion) vol 14 no 3 pp 109ndash118 2015
[11] B Ebrahimi H Bolandhemmat M B Khamesee andF Golnaraghi ldquoA hybrid electromagnetic shock absorber foractive vehicle suspension systemsrdquo Vehicle System Dynamicsvol 49 no 1-2 pp 311ndash332 2011
[12] S Tang L Chen R Wang X Sun and D Shi ldquoResearch onoptimal control of active suspension based on damping multi-modal switchingrdquo Journal of Guangxi University (NaturalScience) vol 39 no 2 pp 300ndash307 2014
[13] R Wang X Ma R Ding X Meng and L Chen ldquoResearchof multi-mode switching control system for hybrid sus-pension based on model referencerdquo Transactions of theChinese Society for Agricultural Machinery vol 48 no 7pp 353ndash360 2017
[14] R Wang Y Qian R Ding X Meng and J Xie ldquoDesign andtests for damping-stiffness of a hybrid electromagnetic sus-pension based on LQGrdquo Journal of Vibration and Shockvol 37 no 3 pp 61ndash65 2017
[15] B Vanavil K K Chaitanya and A S Rao ldquoImproved PIDcontroller design for unstable time delay processes based ondirect synthesis method and maximum sensitivityrdquo Taylorand Francis vol 46 no 8 pp 1349ndash1366 2015
[16] I Mihai and F Andronic ldquoBehavior of a semi-active sus-pension system versus a passive suspension system on anuneven road surfacerdquo Mechanics vol 20 no 1 pp 64ndash692014
[17] D Ngoduy ldquoLinear stability of a generalized multi-anticipative car following model with time delaysrdquo Com-munications in Nonlinear Science and Numerical Simulationvol 22 no 1 pp 420ndash426 2015
[18] H Zhang E Wang F Min R Subash and C Su ldquoSkyhook-based semi-active control of full-vehicle suspension with
magneto-rheological dampersrdquo Chinese Journal of Mechan-ical Engineering vol 26 no 3 pp 498ndash505 2013
[19] F Kou ldquoAn experimental study on the dynamic character-istics of vehicle semi-active seat suspension with magneto-rheological damperrdquo Automotive Engineering vol 37 no 11pp 1346ndash1352 2015
[20] Z Feng S Chen and Y Liang ldquoAn experimental study on thedynamic characteristics of a megneto-rheological semi-activesuspensionrdquoAutomotive Engineering vol 35 no 1 pp 72ndash772013
[21] Z Li and L Xu A New Type of Magnetorheological Damperand Semi-Active coNtrol Design-eory Science Press BeijingChina 2012
[22] B L J Gysen J J H Paulides J L G Janssen andE A Lomonova ldquoActive electromagnetic suspension systemfor improved vehicle dynamicsrdquo IEEE Transactions on Ve-hicular Technology vol 59 no 3 pp 1156ndash1163 2010
[23] Q N Wang S S Liu W H Wang and H Wei ldquoStructuredesign and parameter matching of ball-screw regenerativedamperrdquo Journal of Jilin University (Engineering and Tech-nology Edition) vol 42 no 5 pp 1100ndash1106 2012
[24] S Chen R He and S Lu ldquoEvaluating system of reclaimingenergy suspension comprehensive performancerdquo Trans-actions of the Chinese Society for Agricultural Machineryvol 37 no 7 pp 14ndash18 2006
[25] Y Fan Control Research on Vehicle Suspension with Electro-Hydrostatic Actuator Xirsquoan University of Science and Tech-nology Xirsquoan China 2017
[26] Z Yu Automobile -eory China Machine Press BeijingChina 2009
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Navigation and Observation
International Journal of
Hindawi
wwwhindawicom Volume 2018
Advances in
Multimedia
Submit your manuscripts atwwwhindawicom
where u is induced electromotive force Iz is motor currentand r is internal resistance of the motor
In the active control of the hybrid suspension the in-stantaneous energy consumption power of the ball screwactuator can be expressed as
Pz E middot Iz
ηz (24)
0e instantaneous energy consumption power andconsumption energy can be obtained from formulae(19)sim(24) as follows
Pz _x2 minus _x1 1113857 middot F + FL2πkT 1113857
2middot r1113960 1113961
ηz (25)
Wz 1113946t
0Pz dt (26)
where Pz is motor instantaneous energy consumptionpower ηz is transfer efficiency of ball screw actuator andWzis motor energy consumption
In the energy regeneration of the hybrid suspension theelectromagnetic damping force generated by the ball screwactuator is expressed as
Fs minus2πkT
L1113888 1113889
2
middot_x2 minus _x1 1113857
R + rηz (27)
where Fs is the electromagnetic damping force and R isexternal resistance of the motor
0e instantaneous energy-regenerative power and regen-erative energy of the ball screw actuator can be expressed as
Pk 2πke
L1113888 1113889
2
middot_x2 minus _x1 1113857
2
R + rηz (28)
Wk 1113946t
0Pk dt (29)
where Pk is instantaneous energy-regenerative power andWk is regenerative energy
4 Multimode Coordination Control of HybridActive Suspension
41 -e Active Mode of Damping Switching Control 0esprung mass acceleration is the main evaluation index ofvehicle riding comfort and the dynamic tire load is closelyrelated to vehicle handling and stability An active controlmodel of the hybrid suspension is established by MATLABSimulink software to simulate and analyze the influence ofthe variable damping for the hybrid active suspension onvehicle riding comfort and handling and stability at differentvehicle speeds During the simulation the range of vehiclespeed v is 0ndash120 kmh and the vehicle speed is taken every10 kmh 0e variable damping range of the suspension is200ndash2000 Nmiddotsm and the value of the variable damping istaken every interval 100 Nmiddotsm 0e simulation time is 10 sand the value of r is 05 Ω 0e value of ηz is 097 and thevalue of ηb is 098 0e value of csky is 2000 Nmiddotsm
When the vehicle speed is 30 kmh and 100 kmh re-spectively the RMS of the sprung mass acceleration of thevehicle (aw) and the RMS of the dynamic tire load (DTLrms)change with the variable damping of the hybrid activesuspension as shown in Figures 9 and 10
From Figures 9 and 10 it can be seen that when thevehicle speed is 30 kmh the variable damping values thatmake aw of vehicle and DTLrms minimum are 400 Nmiddotsm and1000 Nmiddotsm respectively and when the vehicle speed is100 kmh the variable damping values that make aw ofvehicle and DTLrms minimum are 500 Nmiddotsm and 1100 Nmiddotsm respectively 0erefore at a certain vehicle speed thevariable damping of the hybrid active suspension cannotmake the best of the vehicle riding comfort and handling andstability at the same time
Suspension performance indexes include the sprung massacceleration suspension working space and dynamic tireload In this paper in order to balance vehicle riding comfortand handling and stability when choosing the variabledamping values of the hybrid suspension in the active controlfor the sprung mass acceleration suspension working spaceand dynamic tire load of the hybrid active suspension thequantitative normalizations and comparative analyses aredone 0at is at the same vehicle speed compared with thepassive suspension the improvement amplitudes of eachperformance index of the hybrid active suspension aremultiplied by different quantification factors and summedAnd the larger the sum the better the dynamic performanceof vehicle Among them the quantification factors of aw theRMS value of suspension working space (SWSrms) andDTLrms are 1 02703 and 01443 respectively [24 25] Andwhen the vehicle speed is 30 kmh and 100 kmh respectivelythe dynamic performance and the active control energyconsumption of the hybrid active suspension change with thevariable damping as shown in Figures 11 and 12
From Figures 11 and 12 when the vehicle speed is 30 kmh the variable damping value of the hybrid active suspensionis 800 Nmiddotsm which makes the vehicle dynamic performancethe best and the active energy consumption the least Whenthe vehicle speed is 100 kmh the variable damping value ofthe hybrid active suspension is 1000 Nmiddotsm which makes thevehicle dynamic performance the best and active energyconsumption the least 0erefore when the vehicle speed is30 kmh and 100 kmh respectively the optimal dampingvalues of the hybrid active suspension are 800 Nmiddotsm and1000 Nmiddotsm respectively When the vehicle speed is 0ndash120 kmh the optimal damping values of the hybrid activesuspension at different vehicle speeds are shown in Figure 13If the damping value which makes the vehicle dynamicperformance the best is different from the damping valuewhich makes the active control energy consumption the leastthe damping value which makes the vehicle dynamic per-formance the best is selected as the optimal damping value ofthe hybrid active suspension at the vehicle speed
When vehicle is in an accelerating or decelerating stateits speed changes rapidly and the range of change is wide sothe vehicle speed value is not easily detected in real time andin order to reduce the energy consumption of the hybridsuspension active controlled and to improve the vehicle
6 Shock and Vibration
riding comfort and handling and stability a variabledamping switching control strategy of the hybrid activesuspension is designed as follows
Fz minus csky middot _x2cs c0 + ck
_v 0
Fz minus csky middot _x2cs c0
_vne 0
(30)
where _v is vehicle acceleration and csky is sky-hook coecient
42 e Semiactive Mode of Feedback Adjustment of Elec-tromagnetic Damping Force MR damper can eectivelyperform semiactive control at ( _x2 minus _x1) _x2 gt 0 so the idealsemiactive control state for hybrid suspension is
Fb minus csky middot _x2 _x2 minus _x1( ) _x2 gt 0
0 _x2 minus _x1( ) _x2 le 0
(31)
From equation (31) the ideal semiactive control force ofthe hybrid suspension minus csky middot _x2 is only related to _x2 when
Vehicle dynamics performanceEnergy consumption of hybrid suspension
Figure 12 Relationship between the variable damping and vehicledynamic performance at 100 kmh speed
20 40 60 80 100 120750
800
850
900
950
1000
1050
Vehicle speed (kmh)
Shoc
k ab
sorb
er d
ampi
ngco
effic
ient
(Nmiddots
m)
Figure 13 Suspension optimal damping at dierent speeds
Shock and Vibration 7
the sky-hook coecient csky is constant However at thistime the ball screw actuator as a power feeding devicegenerates the electromagnetic damping force Fs and acts onthe suspension so that the actual semiactive control force ofthe suspension is dierent from the ideal semiactive controlforce minus csky middot _x2 In this paper the semiactive control modelof the hybrid suspension is established and the changeeects of the dierent output forces on the vehicle ridingcomfort and handling stability are analyzed by MATLABSimulink software e simulation speed is 70 kmh thesimulation time is 5 s and the value of R is 075 Ω edamping comparison of the hybrid suspension in semiactivecontrol is shown in Figure 14
Figure 14 shows that compared with the ideal semi-active control force the actual semiactive control force ofthe hybrid suspension shyuctuates violently and the abso-lute value of the actual semiactive control force is greaterthan the absolute value of the ideal semiactive control force|csky middot _x2| at certain times And a drastic change in theactual semiactive control force makes the suspension notreach ideal semiactive control eect Using electromag-netic damping force feedback adjustment to reduce thedierence between the ideal semiactive control forceminus csky middot _x2 and the actual semiactive control force themethod is as follows
When |csky middot _x2|gt |Fs| the semiactive control force of thehybrid suspension is provided by both the MR damper andthe ball screw actuator and at this point the controllerinputs a controllable current Ik to the MR damper so thatthe Fk output by the MR damper is minus csky middot _x2 minusFs Andwhen |csky middot _x2|le |Fs| the semiactive control force of thehybrid suspension is the Fs which is output by the ball screwactuator and at this point there is no controllable current Ikinput to theMR damper and the function of theMR damperis equivalent to a traditional shock absorber erefore thesemiactive control of the hybrid suspension does not havethe dead zone of traditional electromagnetic semiactivesuspension which helps to improve the semiactive controleect of the hybrid suspension
When there is feedback adjustment the semiactivecontrol force of the hybrid suspension is
Fb Fs Fs
∣∣∣∣∣∣∣∣ge csky middot _x2∣∣∣∣∣
∣∣∣∣∣
minus csky middot _x2 Fs∣∣∣∣∣∣∣∣lt csky middot _x2∣∣∣∣∣
∣∣∣∣∣
(32)
When there is feedback adjustment the Fk output by theMR damper is
Fk 0 Fs
∣∣∣∣∣∣∣∣ge csky middot _x2∣∣∣∣∣
∣∣∣∣∣
minus csky middot _x2 minusFs Fs∣∣∣∣∣∣∣∣lt csky middot _x2∣∣∣∣∣
∣∣∣∣∣
(33)
From equations (31)sim(33) when the electromagneticdamping force feedback adjustment is used the |Fk| outputby theMR damper decreases and when |Fk| decreases it canbe known from equations (8) and (9) that the energyconsumption of the MR damper decreases with it
e comparison of the semiactive control force of thehybrid suspension with or without the electromagneticdamping force feedback adjustment is shown in Figure 15
From Figure 15 the RMS of the ideal semiactive controlforce of the hybrid suspension is 3276 N and when there isno electromagnetic damping force feedback adjustmentthe RMS of the actual semiactive control force of thesuspension is 4041 N and the dierence between the actualsemiactive force of the suspension and the ideal semiactiveforce is 2335 When there is electromagnetic dampingforce feedback adjustment the RMS of the actual semi-active control force of the suspension is 3593 N and thedierence between the actual semiactive force of the sus-pension and the ideal semiactive force is 968 ereforewhen there is electromagnetic damping force feedbackadjustment the actual semiactive control force of thesuspension has a smaller shyuctuation amplitude whichhelps to improve the semiactive control eect of the hybridsuspension
e dynamic responses of the hybrid suspension with orwithout electromagnetic damping force feedback adjust-ment are shown in Figure 16 Among them the damper ofthe passive suspension is the original damper of the vehicleand its damping value is 1600Nmiddotsm
Table 2 shows the response RMS values of the hybridsuspension in semiactive control
From Table 2 compared with the passive suspensionwhen there is electromagnetic damping force feedbackadjustment aw SWSrms and DTLrms of the hybrid sus-pension are reduced by 1698 432 and 1068 re-spectively and compared with the nonfeedback semiactivecontrol when the feedback semiactive control is performedaw SWSrms and DTLrms of the hybrid suspension are re-duced by 252 863 and 671 respectively
From equations (9) and (29) the total system energy ofthe hybrid suspension in semiactive control is
W1 Wk minusWb (34)
whereW1 is the total system energy of the hybrid suspensionin semiactive control
0 1 2 3 4 5ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Hyb
rid su
spen
sion
forc
e (N
)
072 076 080
300
600
Ideal semiactive forceActual semiactive forceElectromagnetic damping force
Figure 14 Damping comparison of the hybrid suspension insemiactive control
8 Shock and Vibration
0 1 2 3 4 5ndash10
ndash5
0
5
10
15
Time (s)
Spru
ng m
ass a
ccel
erat
ion
(ms
2 )
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(a)
0 1 2 3 4 5ndash004
ndash002
000
002
004
006
008
Time (s)
Susp
ensio
n w
orki
ng sp
ace (
m)
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(b)
0 1 2 3 4 5ndash3000
ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Dyn
amic
tire
load
(N)
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(c)
Figure 16 e dynamic responses of the hybrid suspension in semiactive control (a) e response curves of sprung mass acceleration (b)e response curves of suspension working space (c) e response curves of dynamic tire load
0 1 2 3 4 5ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Hyb
rid su
spen
sion
forc
e (N
)
072 076 080
300
600
Ideal semiactive forceWithout feedback semiactive forceWith feedback semiactive force
Figure 15 Semiactive force of the hybrid suspension
Shock and Vibration 9
From equation (33) when the hybrid suspension sem-iactive controlled the curves of the total system energychange over time are shown in Figure 17
From Figure 17 when there is the nonfeedback semiactivecontrol the total system energy of the hybrid suspension is60 J And when there is the feedback semiactive control thetotal system energy of the hybrid suspension is 307 J
43 -e Design of Multimode Coordination ControllerWhen the hybrid suspension is actively controlled thesuspension has good vibration isolation performance buthigh energy consumption And when the hybrid suspensionis semiactively controlled the suspension has good eco-nomic performance but the control has limitations Con-sidering that the ball screw actuator can realize active controlof the hybrid suspension in any suspension state a multi-mode coordinated control strategy of the hybrid suspensionis designed
minus csky middot _x2 Fs1113868111386811138681113868
1113868111386811138681113868gt csky middot _x2
11138681113868111386811138681113868
11138681113868111386811138681113868
⎧⎪⎨
⎪⎩
cs c0
⎧⎪⎪⎪⎨
⎪⎪⎪⎩
_x2 minus _x1 1113857 _x2 gt 0
F Fz minus csky middot _x2
cs c0 + ck1113896 _x2 minus _x1 1113857 _x2 le 0
⎧⎪⎪⎪⎪⎪⎪⎪⎪⎪⎨
⎪⎪⎪⎪⎪⎪⎪⎪⎪⎩
(35)
0e frame diagram of the multimode coordinated controlstrategy of the hybrid suspension is shown in Figure 18
From equations (9) (12) (26) and (29) the total systemenergy of the hybrid suspension in multimode coordinatedcontrol is
W2 Wk minusWb minusWZ minusWc (36)
whereW2 is the total system energy of the hybrid suspensionin multimode coordinated control
A hybrid active suspension simulation model is estab-lished by using MATLABSimulink software From Fig-ure 11 when the hybrid suspension is actively controlled theenergy consumption of the MR damper is mainly affected bythe vehicle speed In order to verify the vibration isolationperformance and energy consumption performance of thehybrid active suspension in the cyclic driving conditionsthis article simulates vehicle urban and suburban conditionswhich is based on GBT 19233-2003 ldquoLight Vehicle FuelConsumption Test Methodrdquo among them the urban con-ditions include four cycle units each cycle time is 195 s thesuburban conditions include one cycle unit and the cycletime is 400 s [26] And the schematic diagram of the urbanconditions unit is shown in Figure 19
In order to simulate the vehicle acceleration signal avehicle speed variation model in different cycle units isestablished by using the signal builder function module inSimulink software On this basis the vehicle accelerationmodel in different cycle units is obtained and the vehicleacceleration model can be used as the switching controlmodel for the variable damping when the hybrid suspensionis actively controlled 0e shock absorbers of active sus-pension and passive suspension adopt the original damper ofvehicle and the damping value c1 is 1600 Nmiddotsm Howeverbecause of the long simulation time the dynamic responsecurves of the hybrid active suspension in urban and sub-urban conditions cannot be displayed Figure 20 shows thedynamic response curve of the hybrid active suspension inthe 145ndash175 s in the urban circulation unit among themvehicle in the 145ndash155 s is in a constant speed vehicle in the155ndash163 s is in the deceleration state and vehicle in the163ndash175 s is in other constant speed
0e dynamic response RMS values of the hybrid sus-pension in urban and suburban circulation units are shownin Tables 3 and 4 respectively
From Figure 20 and Tables 3 and 4 the dynamic re-sponses are good when the hybrid suspension is multimodecoordinated control When in the urban circulation unitcompared with the passive suspension aw SWSrms andDTLrms of the hybrid suspension are reduced by 3943225 and 2081 respectively Compared with the activesuspension aw and DTLrms of the hybrid suspension arereduced by 1322 and 840 respectively but SWSrms isincreased by 805 When in the suburban circulation unitcompared with the passive suspension aw SWSrms andDTLrms of the hybrid suspension are reduced by 3916314 and 1955 respectively Compared with the activesuspension aw and DTLrms of the hybrid suspension arereduced by 1199 and 732 respectively but SWSrms isincreased by 845
0e system energy of the hybrid suspension in urban andsuburban circulation units are shown in Figures 21 and 22respectively
Table 2 0e response root mean square values of the hybridsuspension in semiactive control
Without feedback semiactive controlWith feedback semiactive control
Figure 17 0e system energy of the hybrid suspension in semi-active control
10 Shock and Vibration
en the system energy values of the hybrid suspensionin urban and suburban circulation units are shown inTable 5
From Table 5 when in the urban circulation unit thesystem energy of the active suspension and hybrid sus-pension is minus2165 J and 38 J respectively And when in thesuburban circulation unit the system energy of the activesuspension and hybrid suspension is minus15071 J and minus122 Jrespectively e entire operation cycle includes 4 urbancycle units and 1 suburban cycle unit so the pure energy ofthe active suspension system during the entire operationcycle is minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid suspensionsystem basically realizes energy self-powered in theory
5 Test and Analysis
In order to verify the damping eect of the hybrid sus-pension system in active mode and semiactive mode ahybrid suspension vibration test system is designed as
Suspension controllerSpeed of sprung mass x2
Speed of unsprung mass x1Vehicle acceleration v
Vehicle suspension
Ball screw actuator
Output Fk
Active control force Fz
Ball screw actuator
Output Iz
Output Ik
Electromagnetic damping force Fs
Ideal semiactive control force FL
Output semiactive controlforce Fb = Fs
Output semiactive controlforce Fb = FL = Fk + Fs
x2 ndash x1
(x2 ndash x1) x2 gt 0
(x2 ndash x1) x2 le 0
Output Ik
MR damper
Adjust the duty cycle in real timeaccording to the relationshipbetween active output force
and duty cycle
No controllablecurrent output
MR damper
v ne 0
According to vand ∆v
v = 0
MR damper
Adjustable damping cs = c0 + ck
ndashcsky middot x2
ndashcsky middot x2
|Fs| ge |FL|
|Fs| lt |FL|
Figure 18 e frame diagram of the multimode coordinated control strategy of the hybrid suspension
Vehicle shifting
00
15
10
30
30
20
45
40
50
60
60
75 90 105 120Time (s)
Veh
icle
spee
d (k
mh
)
135 150 165 180 195
Figure 19 e schematic diagram of the urban conditions unit
Shock and Vibration 11
shown in Figure 23 During the test the MR damper has nocontrollable current input in the active control mode of thehybrid suspension and its damping value is always 800Nmiddotsm
e passive suspension semiactive suspension and activesuspension all adopt the original damper of the vehicle andthe damping value c1 is 1600Nmiddotsm
Figure 20e dynamic responses of the hybrid suspension (a)e response curves of sprungmass acceleration (b)e response curves ofsuspension working space (c) e response curves of dynamic tire load
12 Shock and Vibration
Because of the limitation of test conditions only thesprung mass acceleration dynamic response of the hybridsuspension is measured in this test e dynamic response ofthe sprung mass acceleration of the hybrid suspension inactive mode under random road is shown in Figure 24 epower spectrum of the sprung mass acceleration of thehybrid suspension in active mode is shown in Figure 25
Table 6 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in active mode
From Table 6 compared with passive suspension aw ofthe hybrid suspension in active mode is reduced by 3945and compared with active suspension aw of the hybridsuspension in active mode is reduced by 1432 And fromFigure 25 when the hybrid suspension is actively controlledcompared with passive suspension the vibration isolationperformance of the suspension in the low frequency andlow-frequency resonance areas is similar to that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withactive suspension the vibration isolation performance of thesuspension in the low frequency high frequency and res-onance regions is slightly better
e dynamic response of the sprung mass acceleration ofthe hybrid suspension in semiactive mode under randomroad is shown in Figure 26 e power spectrum of thesprung mass acceleration of the hybrid suspension insemiactive mode is shown in Figure 27
Table 7 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in semiactive mode
From Table 7 compared with passive suspension aw ofthe hybrid suspension in semiactive mode is reduced by1642 Compared with semiactive suspension aw of thehybrid suspension in active mode is reduced by 307 Andfrom Figure 27 when the hybrid suspension is semiactivelycontrolled compared with passive suspension the vibration
Table 3 e dynamic response root mean square values of thehybrid suspension in urban circulation unit
Figure 21 e system energy of the hybrid suspension in urbancirculation units
0 100 200 300 400ndash20000
ndash15000
ndash10000
ndash5000
0
5000
Time (s)
Susp
ensio
n sy
stem
ener
gy (J
)
Active suspensionHybrid suspension
Figure 22e system energy of the hybrid suspension in suburbancirculation units
Table 5 e system energy values of the hybrid suspension (J)
Cycle unit Active suspension Hybrid suspensionUrban minus2165 38Suburban minus15071 minus122
Figure 23 e test system of the hybrid active suspension
Shock and Vibration 13
isolation performance of the suspension in the low frequencyand low-frequency resonance areas is worse than that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withsemiactive suspension the vibration isolation performance ofthe suspension in the low frequency high frequency andresonance regions is slightly better
e energy consumption power and energy-regenerativepower of the hybrid suspension are shown in Figure 28
From Figure 28 in the active mode the average powerconsumption of the ball screw actuator is 2561W In thesemiactive mode the average regenerative power of the ballscrew actuator is 2696W and the average of theMR damperenergy consumption power is 178W erefore the pureaverage power consumption of the hybrid suspension isminus043W and the test results are basically consistent with thesimulation
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spr
ung
mas
s acc
eler
atio
n (m
s2 )
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 24 e dynamic response of the sprung mass accelerationof the hybrid suspension in active mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 25 e power spectrum of the sprung mass acceleration ofthe hybrid suspension in active mode
Table 6 e sprung mass acceleration RMS of the hybrid sus-pension in active mode
Indicators Passivesuspension
Activesuspension
Hybridsuspension
aw (ms2) 13378 09423 08074
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spru
ng m
ass a
ccel
erat
ion
(ms
2 )
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 26 e dynamic responses of the sprung mass accelerationof the hybrid suspension in semiactive mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 27e power spectrums of the sprungmass acceleration ofthe hybrid suspension in semiactive mode
Table 7 e sprung mass acceleration RMS of the hybrid sus-pension in semiactive mode
Indicators Passivesuspension
Semiactivesuspension
Hybridsuspension
aw (ms2) 13378 11536 11182
14 Shock and Vibration
6 Conclusion
(1) A ball screw actuator andMR damper are introducedinto vehicle suspension system and a new kind ofhybrid active suspension structure is put forwarde ball screw actuator is prototyped and thefunctional relationship between the back-EMF co-ecient the electromagnetic torque coecient of themotor and the suspension vibration speed is ob-tained by test analyses And the active output me-chanical properties of the ball screw actuator aretested and the results show that the actuator hasgood active output force characteristics
(2) e inshyuences of the variable damping value of thesuspension system on the riding comfort handlingand stability and energy consumption characteristicsof the hybrid suspension in the active control modeare analyzeden the optimal damping values of thehybrid suspension at dierent vehicle speeds aredesigned e eects of electromagnetic dampingforce on the actual semiactive force and the systemenergy of the suspension in the semiactive controlmode are analyzed and then the hybrid suspensionwith semiactive mode which has electromagneticdamping force feedback adjustment is designed Onthis basis a multimode coordinated control strategyfor the hybrid suspension is designed
(3) e damping performance and energy consumptioncharacteristics of the hybrid suspension under cyclicdriving condition are simulated by MATLABSimulink software and the results show that whenin the urban circulation unit compared with activesuspension aw and DTLrms of the hybrid suspensionare reduced by 1322 and 840 respectively Andwhen in the suburban circulation unit comparedwith the active suspension aw and DTLrms of thehybrid suspension are reduced by 1199 and 732respectively e pure energy of the active suspen-sion system during the entire operation cycle is
minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid sus-pension system basically realizes energy self-poweredin theory
(4) e eectiveness verication test of the hybrid sus-pension in active mode and semiactive mode controlis carried out and the results show that when thehybrid suspension is actively controlled comparedwith active suspension aw of the hybrid suspension inactive mode is reduced by 1432 When the hybridsuspension is semiactively controlled compared withsemiactive suspension aw of the hybrid suspension inactive mode is reduced by 307 e test and sim-ulation results are basically consistent and the testveries the correctness of the simulation
Data Availability
e data used to support the ndings of this study are in-cluded within the supplementary information les And thedata used to support the ndings of this study are availablefrom the corresponding author upon request
Conflicts of Interest
e authors declare that there are no conshyicts of interestregarding the publication of this paper
Acknowledgments
is work was supported by the National Natural ScienceFoundation of China (Grant no 51775426) Service LocalSpecial Program Support Project of Shaanxi ProvincialEducation Department (Grant no 17JF017) and XirsquoanScience and Technology Program Funding Project (Grantno 2017079CGRC042-XAKD007)
Supplementary Materials
e supplementary materials are this articlersquos experimentaldata including (1) MR damperrsquos characteristic test data (2)the motor counter electromotive forcersquos peak test data (3)the ball screw actuatorrsquos active output force test data (4) thehybrid suspensionrsquos sprung mass acceleration test data inactive mode (5) the hybrid suspensionrsquos sprung mass ac-celeration test data in semiactive mode and (6) the hybridsuspensionrsquos energy consumption power and energy re-generation power test data (Supplementary Materials)
References
[1] M Montazeri-Gh and O Kavianipour ldquoInvestigation of theactive electromagnetic suspension system considering hybridcontrol strategyrdquo Proceedings of the Institution of MechanicalEngineers Part C Journal of Mechanical Engineering Sciencevol 228 no 10 pp 1658ndash1669 2013
[2] H E Tseng and D Hrovat ldquoState of the art survey active andsemi-active suspension controlrdquo Vehicle System Dynamicsvol 53 no 7 pp 1034ndash1062 2015
0 1 2 3 4 50
20
40
60
80
Time (s)
Pow
er (W
)
Energy consumption power in active modeEnergy regenerative power in semiactive modeEnergy consumption power in semiactive mode
Figure 28 e energy consumption power and energy-re-generative power of the hybrid suspension
Shock and Vibration 15
[3] F Kou J Du Z Wang D Li and J Xu ldquoNonlinear modelingand coordinate optimization of a semi-active energy re-generative suspension with an electro-hydraulic actuatorrdquoChina Mechanical Engineering vol 28 no 14 pp 1701ndash17072017
[4] S H Zareh A Sarrafan A A A Khayyat and A ZabihollahldquoIntelligent semi-active vibration control of eleven degrees offreedom suspension system using magnetorheologicaldampersrdquo Journal of Mechanical Science and Technologyvol 26 no 3 pp 323ndash334 2012
[5] S A Chen X Li L J Zhao Y X Wang and Y B KimldquoDevelopment of a control method for an electromagneticsemi-active suspension reclaiming energy with varying chargevoltage in stepsrdquo International Journal of Automotive Tech-nology vol 16 no 5 pp 765ndash773 2015
[6] K Nakano Y Suda and S Nakadai ldquoSelf-powered activevibration control using a single electric actuatorrdquo Journal ofSound and Vibration vol 260 no 2 pp 213ndash235 2003
[7] K Nakano ldquoCombined type self-powered active vibrationcontrol of truck cabinsrdquo Vehicle System Dynamics vol 41no 6 pp 449ndash473 2004
[8] K Huang Y C Zhang F Yu and Y H Gu ldquoCoordinateoptimization for synthetical performance of electrical energy-regenerative active suspensionrdquo Journal of Shanghai JiaotongUniversity vol 43 no 2 pp 226ndash230 2009
[9] K Huang F Yu and Y C Zhang ldquoActive control of energy-regenerative electromagnetic suspension based on energy flowanalysisrdquo Journal of Shanghai Jiaotong University vol 45no 67 pp 1068ndash1073 2011
[10] D S Huang J Q Zhang Y L Liu L Yi and X Y WangldquoPerformance of a novel energy-regenerative active suspen-sion systemrdquo Journal of Chongqing University (English Edi-tion) vol 14 no 3 pp 109ndash118 2015
[11] B Ebrahimi H Bolandhemmat M B Khamesee andF Golnaraghi ldquoA hybrid electromagnetic shock absorber foractive vehicle suspension systemsrdquo Vehicle System Dynamicsvol 49 no 1-2 pp 311ndash332 2011
[12] S Tang L Chen R Wang X Sun and D Shi ldquoResearch onoptimal control of active suspension based on damping multi-modal switchingrdquo Journal of Guangxi University (NaturalScience) vol 39 no 2 pp 300ndash307 2014
[13] R Wang X Ma R Ding X Meng and L Chen ldquoResearchof multi-mode switching control system for hybrid sus-pension based on model referencerdquo Transactions of theChinese Society for Agricultural Machinery vol 48 no 7pp 353ndash360 2017
[14] R Wang Y Qian R Ding X Meng and J Xie ldquoDesign andtests for damping-stiffness of a hybrid electromagnetic sus-pension based on LQGrdquo Journal of Vibration and Shockvol 37 no 3 pp 61ndash65 2017
[15] B Vanavil K K Chaitanya and A S Rao ldquoImproved PIDcontroller design for unstable time delay processes based ondirect synthesis method and maximum sensitivityrdquo Taylorand Francis vol 46 no 8 pp 1349ndash1366 2015
[16] I Mihai and F Andronic ldquoBehavior of a semi-active sus-pension system versus a passive suspension system on anuneven road surfacerdquo Mechanics vol 20 no 1 pp 64ndash692014
[17] D Ngoduy ldquoLinear stability of a generalized multi-anticipative car following model with time delaysrdquo Com-munications in Nonlinear Science and Numerical Simulationvol 22 no 1 pp 420ndash426 2015
[18] H Zhang E Wang F Min R Subash and C Su ldquoSkyhook-based semi-active control of full-vehicle suspension with
magneto-rheological dampersrdquo Chinese Journal of Mechan-ical Engineering vol 26 no 3 pp 498ndash505 2013
[19] F Kou ldquoAn experimental study on the dynamic character-istics of vehicle semi-active seat suspension with magneto-rheological damperrdquo Automotive Engineering vol 37 no 11pp 1346ndash1352 2015
[20] Z Feng S Chen and Y Liang ldquoAn experimental study on thedynamic characteristics of a megneto-rheological semi-activesuspensionrdquoAutomotive Engineering vol 35 no 1 pp 72ndash772013
[21] Z Li and L Xu A New Type of Magnetorheological Damperand Semi-Active coNtrol Design-eory Science Press BeijingChina 2012
[22] B L J Gysen J J H Paulides J L G Janssen andE A Lomonova ldquoActive electromagnetic suspension systemfor improved vehicle dynamicsrdquo IEEE Transactions on Ve-hicular Technology vol 59 no 3 pp 1156ndash1163 2010
[23] Q N Wang S S Liu W H Wang and H Wei ldquoStructuredesign and parameter matching of ball-screw regenerativedamperrdquo Journal of Jilin University (Engineering and Tech-nology Edition) vol 42 no 5 pp 1100ndash1106 2012
[24] S Chen R He and S Lu ldquoEvaluating system of reclaimingenergy suspension comprehensive performancerdquo Trans-actions of the Chinese Society for Agricultural Machineryvol 37 no 7 pp 14ndash18 2006
[25] Y Fan Control Research on Vehicle Suspension with Electro-Hydrostatic Actuator Xirsquoan University of Science and Tech-nology Xirsquoan China 2017
[26] Z Yu Automobile -eory China Machine Press BeijingChina 2009
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Navigation and Observation
International Journal of
Hindawi
wwwhindawicom Volume 2018
Advances in
Multimedia
Submit your manuscripts atwwwhindawicom
riding comfort and handling and stability a variabledamping switching control strategy of the hybrid activesuspension is designed as follows
Fz minus csky middot _x2cs c0 + ck
_v 0
Fz minus csky middot _x2cs c0
_vne 0
(30)
where _v is vehicle acceleration and csky is sky-hook coecient
42 e Semiactive Mode of Feedback Adjustment of Elec-tromagnetic Damping Force MR damper can eectivelyperform semiactive control at ( _x2 minus _x1) _x2 gt 0 so the idealsemiactive control state for hybrid suspension is
Fb minus csky middot _x2 _x2 minus _x1( ) _x2 gt 0
0 _x2 minus _x1( ) _x2 le 0
(31)
From equation (31) the ideal semiactive control force ofthe hybrid suspension minus csky middot _x2 is only related to _x2 when
Vehicle dynamics performanceEnergy consumption of hybrid suspension
Figure 12 Relationship between the variable damping and vehicledynamic performance at 100 kmh speed
20 40 60 80 100 120750
800
850
900
950
1000
1050
Vehicle speed (kmh)
Shoc
k ab
sorb
er d
ampi
ngco
effic
ient
(Nmiddots
m)
Figure 13 Suspension optimal damping at dierent speeds
Shock and Vibration 7
the sky-hook coecient csky is constant However at thistime the ball screw actuator as a power feeding devicegenerates the electromagnetic damping force Fs and acts onthe suspension so that the actual semiactive control force ofthe suspension is dierent from the ideal semiactive controlforce minus csky middot _x2 In this paper the semiactive control modelof the hybrid suspension is established and the changeeects of the dierent output forces on the vehicle ridingcomfort and handling stability are analyzed by MATLABSimulink software e simulation speed is 70 kmh thesimulation time is 5 s and the value of R is 075 Ω edamping comparison of the hybrid suspension in semiactivecontrol is shown in Figure 14
Figure 14 shows that compared with the ideal semi-active control force the actual semiactive control force ofthe hybrid suspension shyuctuates violently and the abso-lute value of the actual semiactive control force is greaterthan the absolute value of the ideal semiactive control force|csky middot _x2| at certain times And a drastic change in theactual semiactive control force makes the suspension notreach ideal semiactive control eect Using electromag-netic damping force feedback adjustment to reduce thedierence between the ideal semiactive control forceminus csky middot _x2 and the actual semiactive control force themethod is as follows
When |csky middot _x2|gt |Fs| the semiactive control force of thehybrid suspension is provided by both the MR damper andthe ball screw actuator and at this point the controllerinputs a controllable current Ik to the MR damper so thatthe Fk output by the MR damper is minus csky middot _x2 minusFs Andwhen |csky middot _x2|le |Fs| the semiactive control force of thehybrid suspension is the Fs which is output by the ball screwactuator and at this point there is no controllable current Ikinput to theMR damper and the function of theMR damperis equivalent to a traditional shock absorber erefore thesemiactive control of the hybrid suspension does not havethe dead zone of traditional electromagnetic semiactivesuspension which helps to improve the semiactive controleect of the hybrid suspension
When there is feedback adjustment the semiactivecontrol force of the hybrid suspension is
Fb Fs Fs
∣∣∣∣∣∣∣∣ge csky middot _x2∣∣∣∣∣
∣∣∣∣∣
minus csky middot _x2 Fs∣∣∣∣∣∣∣∣lt csky middot _x2∣∣∣∣∣
∣∣∣∣∣
(32)
When there is feedback adjustment the Fk output by theMR damper is
Fk 0 Fs
∣∣∣∣∣∣∣∣ge csky middot _x2∣∣∣∣∣
∣∣∣∣∣
minus csky middot _x2 minusFs Fs∣∣∣∣∣∣∣∣lt csky middot _x2∣∣∣∣∣
∣∣∣∣∣
(33)
From equations (31)sim(33) when the electromagneticdamping force feedback adjustment is used the |Fk| outputby theMR damper decreases and when |Fk| decreases it canbe known from equations (8) and (9) that the energyconsumption of the MR damper decreases with it
e comparison of the semiactive control force of thehybrid suspension with or without the electromagneticdamping force feedback adjustment is shown in Figure 15
From Figure 15 the RMS of the ideal semiactive controlforce of the hybrid suspension is 3276 N and when there isno electromagnetic damping force feedback adjustmentthe RMS of the actual semiactive control force of thesuspension is 4041 N and the dierence between the actualsemiactive force of the suspension and the ideal semiactiveforce is 2335 When there is electromagnetic dampingforce feedback adjustment the RMS of the actual semi-active control force of the suspension is 3593 N and thedierence between the actual semiactive force of the sus-pension and the ideal semiactive force is 968 ereforewhen there is electromagnetic damping force feedbackadjustment the actual semiactive control force of thesuspension has a smaller shyuctuation amplitude whichhelps to improve the semiactive control eect of the hybridsuspension
e dynamic responses of the hybrid suspension with orwithout electromagnetic damping force feedback adjust-ment are shown in Figure 16 Among them the damper ofthe passive suspension is the original damper of the vehicleand its damping value is 1600Nmiddotsm
Table 2 shows the response RMS values of the hybridsuspension in semiactive control
From Table 2 compared with the passive suspensionwhen there is electromagnetic damping force feedbackadjustment aw SWSrms and DTLrms of the hybrid sus-pension are reduced by 1698 432 and 1068 re-spectively and compared with the nonfeedback semiactivecontrol when the feedback semiactive control is performedaw SWSrms and DTLrms of the hybrid suspension are re-duced by 252 863 and 671 respectively
From equations (9) and (29) the total system energy ofthe hybrid suspension in semiactive control is
W1 Wk minusWb (34)
whereW1 is the total system energy of the hybrid suspensionin semiactive control
0 1 2 3 4 5ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Hyb
rid su
spen
sion
forc
e (N
)
072 076 080
300
600
Ideal semiactive forceActual semiactive forceElectromagnetic damping force
Figure 14 Damping comparison of the hybrid suspension insemiactive control
8 Shock and Vibration
0 1 2 3 4 5ndash10
ndash5
0
5
10
15
Time (s)
Spru
ng m
ass a
ccel
erat
ion
(ms
2 )
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(a)
0 1 2 3 4 5ndash004
ndash002
000
002
004
006
008
Time (s)
Susp
ensio
n w
orki
ng sp
ace (
m)
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(b)
0 1 2 3 4 5ndash3000
ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Dyn
amic
tire
load
(N)
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(c)
Figure 16 e dynamic responses of the hybrid suspension in semiactive control (a) e response curves of sprung mass acceleration (b)e response curves of suspension working space (c) e response curves of dynamic tire load
0 1 2 3 4 5ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Hyb
rid su
spen
sion
forc
e (N
)
072 076 080
300
600
Ideal semiactive forceWithout feedback semiactive forceWith feedback semiactive force
Figure 15 Semiactive force of the hybrid suspension
Shock and Vibration 9
From equation (33) when the hybrid suspension sem-iactive controlled the curves of the total system energychange over time are shown in Figure 17
From Figure 17 when there is the nonfeedback semiactivecontrol the total system energy of the hybrid suspension is60 J And when there is the feedback semiactive control thetotal system energy of the hybrid suspension is 307 J
43 -e Design of Multimode Coordination ControllerWhen the hybrid suspension is actively controlled thesuspension has good vibration isolation performance buthigh energy consumption And when the hybrid suspensionis semiactively controlled the suspension has good eco-nomic performance but the control has limitations Con-sidering that the ball screw actuator can realize active controlof the hybrid suspension in any suspension state a multi-mode coordinated control strategy of the hybrid suspensionis designed
minus csky middot _x2 Fs1113868111386811138681113868
1113868111386811138681113868gt csky middot _x2
11138681113868111386811138681113868
11138681113868111386811138681113868
⎧⎪⎨
⎪⎩
cs c0
⎧⎪⎪⎪⎨
⎪⎪⎪⎩
_x2 minus _x1 1113857 _x2 gt 0
F Fz minus csky middot _x2
cs c0 + ck1113896 _x2 minus _x1 1113857 _x2 le 0
⎧⎪⎪⎪⎪⎪⎪⎪⎪⎪⎨
⎪⎪⎪⎪⎪⎪⎪⎪⎪⎩
(35)
0e frame diagram of the multimode coordinated controlstrategy of the hybrid suspension is shown in Figure 18
From equations (9) (12) (26) and (29) the total systemenergy of the hybrid suspension in multimode coordinatedcontrol is
W2 Wk minusWb minusWZ minusWc (36)
whereW2 is the total system energy of the hybrid suspensionin multimode coordinated control
A hybrid active suspension simulation model is estab-lished by using MATLABSimulink software From Fig-ure 11 when the hybrid suspension is actively controlled theenergy consumption of the MR damper is mainly affected bythe vehicle speed In order to verify the vibration isolationperformance and energy consumption performance of thehybrid active suspension in the cyclic driving conditionsthis article simulates vehicle urban and suburban conditionswhich is based on GBT 19233-2003 ldquoLight Vehicle FuelConsumption Test Methodrdquo among them the urban con-ditions include four cycle units each cycle time is 195 s thesuburban conditions include one cycle unit and the cycletime is 400 s [26] And the schematic diagram of the urbanconditions unit is shown in Figure 19
In order to simulate the vehicle acceleration signal avehicle speed variation model in different cycle units isestablished by using the signal builder function module inSimulink software On this basis the vehicle accelerationmodel in different cycle units is obtained and the vehicleacceleration model can be used as the switching controlmodel for the variable damping when the hybrid suspensionis actively controlled 0e shock absorbers of active sus-pension and passive suspension adopt the original damper ofvehicle and the damping value c1 is 1600 Nmiddotsm Howeverbecause of the long simulation time the dynamic responsecurves of the hybrid active suspension in urban and sub-urban conditions cannot be displayed Figure 20 shows thedynamic response curve of the hybrid active suspension inthe 145ndash175 s in the urban circulation unit among themvehicle in the 145ndash155 s is in a constant speed vehicle in the155ndash163 s is in the deceleration state and vehicle in the163ndash175 s is in other constant speed
0e dynamic response RMS values of the hybrid sus-pension in urban and suburban circulation units are shownin Tables 3 and 4 respectively
From Figure 20 and Tables 3 and 4 the dynamic re-sponses are good when the hybrid suspension is multimodecoordinated control When in the urban circulation unitcompared with the passive suspension aw SWSrms andDTLrms of the hybrid suspension are reduced by 3943225 and 2081 respectively Compared with the activesuspension aw and DTLrms of the hybrid suspension arereduced by 1322 and 840 respectively but SWSrms isincreased by 805 When in the suburban circulation unitcompared with the passive suspension aw SWSrms andDTLrms of the hybrid suspension are reduced by 3916314 and 1955 respectively Compared with the activesuspension aw and DTLrms of the hybrid suspension arereduced by 1199 and 732 respectively but SWSrms isincreased by 845
0e system energy of the hybrid suspension in urban andsuburban circulation units are shown in Figures 21 and 22respectively
Table 2 0e response root mean square values of the hybridsuspension in semiactive control
Without feedback semiactive controlWith feedback semiactive control
Figure 17 0e system energy of the hybrid suspension in semi-active control
10 Shock and Vibration
en the system energy values of the hybrid suspensionin urban and suburban circulation units are shown inTable 5
From Table 5 when in the urban circulation unit thesystem energy of the active suspension and hybrid sus-pension is minus2165 J and 38 J respectively And when in thesuburban circulation unit the system energy of the activesuspension and hybrid suspension is minus15071 J and minus122 Jrespectively e entire operation cycle includes 4 urbancycle units and 1 suburban cycle unit so the pure energy ofthe active suspension system during the entire operationcycle is minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid suspensionsystem basically realizes energy self-powered in theory
5 Test and Analysis
In order to verify the damping eect of the hybrid sus-pension system in active mode and semiactive mode ahybrid suspension vibration test system is designed as
Suspension controllerSpeed of sprung mass x2
Speed of unsprung mass x1Vehicle acceleration v
Vehicle suspension
Ball screw actuator
Output Fk
Active control force Fz
Ball screw actuator
Output Iz
Output Ik
Electromagnetic damping force Fs
Ideal semiactive control force FL
Output semiactive controlforce Fb = Fs
Output semiactive controlforce Fb = FL = Fk + Fs
x2 ndash x1
(x2 ndash x1) x2 gt 0
(x2 ndash x1) x2 le 0
Output Ik
MR damper
Adjust the duty cycle in real timeaccording to the relationshipbetween active output force
and duty cycle
No controllablecurrent output
MR damper
v ne 0
According to vand ∆v
v = 0
MR damper
Adjustable damping cs = c0 + ck
ndashcsky middot x2
ndashcsky middot x2
|Fs| ge |FL|
|Fs| lt |FL|
Figure 18 e frame diagram of the multimode coordinated control strategy of the hybrid suspension
Vehicle shifting
00
15
10
30
30
20
45
40
50
60
60
75 90 105 120Time (s)
Veh
icle
spee
d (k
mh
)
135 150 165 180 195
Figure 19 e schematic diagram of the urban conditions unit
Shock and Vibration 11
shown in Figure 23 During the test the MR damper has nocontrollable current input in the active control mode of thehybrid suspension and its damping value is always 800Nmiddotsm
e passive suspension semiactive suspension and activesuspension all adopt the original damper of the vehicle andthe damping value c1 is 1600Nmiddotsm
Figure 20e dynamic responses of the hybrid suspension (a)e response curves of sprungmass acceleration (b)e response curves ofsuspension working space (c) e response curves of dynamic tire load
12 Shock and Vibration
Because of the limitation of test conditions only thesprung mass acceleration dynamic response of the hybridsuspension is measured in this test e dynamic response ofthe sprung mass acceleration of the hybrid suspension inactive mode under random road is shown in Figure 24 epower spectrum of the sprung mass acceleration of thehybrid suspension in active mode is shown in Figure 25
Table 6 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in active mode
From Table 6 compared with passive suspension aw ofthe hybrid suspension in active mode is reduced by 3945and compared with active suspension aw of the hybridsuspension in active mode is reduced by 1432 And fromFigure 25 when the hybrid suspension is actively controlledcompared with passive suspension the vibration isolationperformance of the suspension in the low frequency andlow-frequency resonance areas is similar to that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withactive suspension the vibration isolation performance of thesuspension in the low frequency high frequency and res-onance regions is slightly better
e dynamic response of the sprung mass acceleration ofthe hybrid suspension in semiactive mode under randomroad is shown in Figure 26 e power spectrum of thesprung mass acceleration of the hybrid suspension insemiactive mode is shown in Figure 27
Table 7 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in semiactive mode
From Table 7 compared with passive suspension aw ofthe hybrid suspension in semiactive mode is reduced by1642 Compared with semiactive suspension aw of thehybrid suspension in active mode is reduced by 307 Andfrom Figure 27 when the hybrid suspension is semiactivelycontrolled compared with passive suspension the vibration
Table 3 e dynamic response root mean square values of thehybrid suspension in urban circulation unit
Figure 21 e system energy of the hybrid suspension in urbancirculation units
0 100 200 300 400ndash20000
ndash15000
ndash10000
ndash5000
0
5000
Time (s)
Susp
ensio
n sy
stem
ener
gy (J
)
Active suspensionHybrid suspension
Figure 22e system energy of the hybrid suspension in suburbancirculation units
Table 5 e system energy values of the hybrid suspension (J)
Cycle unit Active suspension Hybrid suspensionUrban minus2165 38Suburban minus15071 minus122
Figure 23 e test system of the hybrid active suspension
Shock and Vibration 13
isolation performance of the suspension in the low frequencyand low-frequency resonance areas is worse than that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withsemiactive suspension the vibration isolation performance ofthe suspension in the low frequency high frequency andresonance regions is slightly better
e energy consumption power and energy-regenerativepower of the hybrid suspension are shown in Figure 28
From Figure 28 in the active mode the average powerconsumption of the ball screw actuator is 2561W In thesemiactive mode the average regenerative power of the ballscrew actuator is 2696W and the average of theMR damperenergy consumption power is 178W erefore the pureaverage power consumption of the hybrid suspension isminus043W and the test results are basically consistent with thesimulation
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spr
ung
mas
s acc
eler
atio
n (m
s2 )
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 24 e dynamic response of the sprung mass accelerationof the hybrid suspension in active mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 25 e power spectrum of the sprung mass acceleration ofthe hybrid suspension in active mode
Table 6 e sprung mass acceleration RMS of the hybrid sus-pension in active mode
Indicators Passivesuspension
Activesuspension
Hybridsuspension
aw (ms2) 13378 09423 08074
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spru
ng m
ass a
ccel
erat
ion
(ms
2 )
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 26 e dynamic responses of the sprung mass accelerationof the hybrid suspension in semiactive mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 27e power spectrums of the sprungmass acceleration ofthe hybrid suspension in semiactive mode
Table 7 e sprung mass acceleration RMS of the hybrid sus-pension in semiactive mode
Indicators Passivesuspension
Semiactivesuspension
Hybridsuspension
aw (ms2) 13378 11536 11182
14 Shock and Vibration
6 Conclusion
(1) A ball screw actuator andMR damper are introducedinto vehicle suspension system and a new kind ofhybrid active suspension structure is put forwarde ball screw actuator is prototyped and thefunctional relationship between the back-EMF co-ecient the electromagnetic torque coecient of themotor and the suspension vibration speed is ob-tained by test analyses And the active output me-chanical properties of the ball screw actuator aretested and the results show that the actuator hasgood active output force characteristics
(2) e inshyuences of the variable damping value of thesuspension system on the riding comfort handlingand stability and energy consumption characteristicsof the hybrid suspension in the active control modeare analyzeden the optimal damping values of thehybrid suspension at dierent vehicle speeds aredesigned e eects of electromagnetic dampingforce on the actual semiactive force and the systemenergy of the suspension in the semiactive controlmode are analyzed and then the hybrid suspensionwith semiactive mode which has electromagneticdamping force feedback adjustment is designed Onthis basis a multimode coordinated control strategyfor the hybrid suspension is designed
(3) e damping performance and energy consumptioncharacteristics of the hybrid suspension under cyclicdriving condition are simulated by MATLABSimulink software and the results show that whenin the urban circulation unit compared with activesuspension aw and DTLrms of the hybrid suspensionare reduced by 1322 and 840 respectively Andwhen in the suburban circulation unit comparedwith the active suspension aw and DTLrms of thehybrid suspension are reduced by 1199 and 732respectively e pure energy of the active suspen-sion system during the entire operation cycle is
minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid sus-pension system basically realizes energy self-poweredin theory
(4) e eectiveness verication test of the hybrid sus-pension in active mode and semiactive mode controlis carried out and the results show that when thehybrid suspension is actively controlled comparedwith active suspension aw of the hybrid suspension inactive mode is reduced by 1432 When the hybridsuspension is semiactively controlled compared withsemiactive suspension aw of the hybrid suspension inactive mode is reduced by 307 e test and sim-ulation results are basically consistent and the testveries the correctness of the simulation
Data Availability
e data used to support the ndings of this study are in-cluded within the supplementary information les And thedata used to support the ndings of this study are availablefrom the corresponding author upon request
Conflicts of Interest
e authors declare that there are no conshyicts of interestregarding the publication of this paper
Acknowledgments
is work was supported by the National Natural ScienceFoundation of China (Grant no 51775426) Service LocalSpecial Program Support Project of Shaanxi ProvincialEducation Department (Grant no 17JF017) and XirsquoanScience and Technology Program Funding Project (Grantno 2017079CGRC042-XAKD007)
Supplementary Materials
e supplementary materials are this articlersquos experimentaldata including (1) MR damperrsquos characteristic test data (2)the motor counter electromotive forcersquos peak test data (3)the ball screw actuatorrsquos active output force test data (4) thehybrid suspensionrsquos sprung mass acceleration test data inactive mode (5) the hybrid suspensionrsquos sprung mass ac-celeration test data in semiactive mode and (6) the hybridsuspensionrsquos energy consumption power and energy re-generation power test data (Supplementary Materials)
References
[1] M Montazeri-Gh and O Kavianipour ldquoInvestigation of theactive electromagnetic suspension system considering hybridcontrol strategyrdquo Proceedings of the Institution of MechanicalEngineers Part C Journal of Mechanical Engineering Sciencevol 228 no 10 pp 1658ndash1669 2013
[2] H E Tseng and D Hrovat ldquoState of the art survey active andsemi-active suspension controlrdquo Vehicle System Dynamicsvol 53 no 7 pp 1034ndash1062 2015
0 1 2 3 4 50
20
40
60
80
Time (s)
Pow
er (W
)
Energy consumption power in active modeEnergy regenerative power in semiactive modeEnergy consumption power in semiactive mode
Figure 28 e energy consumption power and energy-re-generative power of the hybrid suspension
Shock and Vibration 15
[3] F Kou J Du Z Wang D Li and J Xu ldquoNonlinear modelingand coordinate optimization of a semi-active energy re-generative suspension with an electro-hydraulic actuatorrdquoChina Mechanical Engineering vol 28 no 14 pp 1701ndash17072017
[4] S H Zareh A Sarrafan A A A Khayyat and A ZabihollahldquoIntelligent semi-active vibration control of eleven degrees offreedom suspension system using magnetorheologicaldampersrdquo Journal of Mechanical Science and Technologyvol 26 no 3 pp 323ndash334 2012
[5] S A Chen X Li L J Zhao Y X Wang and Y B KimldquoDevelopment of a control method for an electromagneticsemi-active suspension reclaiming energy with varying chargevoltage in stepsrdquo International Journal of Automotive Tech-nology vol 16 no 5 pp 765ndash773 2015
[6] K Nakano Y Suda and S Nakadai ldquoSelf-powered activevibration control using a single electric actuatorrdquo Journal ofSound and Vibration vol 260 no 2 pp 213ndash235 2003
[7] K Nakano ldquoCombined type self-powered active vibrationcontrol of truck cabinsrdquo Vehicle System Dynamics vol 41no 6 pp 449ndash473 2004
[8] K Huang Y C Zhang F Yu and Y H Gu ldquoCoordinateoptimization for synthetical performance of electrical energy-regenerative active suspensionrdquo Journal of Shanghai JiaotongUniversity vol 43 no 2 pp 226ndash230 2009
[9] K Huang F Yu and Y C Zhang ldquoActive control of energy-regenerative electromagnetic suspension based on energy flowanalysisrdquo Journal of Shanghai Jiaotong University vol 45no 67 pp 1068ndash1073 2011
[10] D S Huang J Q Zhang Y L Liu L Yi and X Y WangldquoPerformance of a novel energy-regenerative active suspen-sion systemrdquo Journal of Chongqing University (English Edi-tion) vol 14 no 3 pp 109ndash118 2015
[11] B Ebrahimi H Bolandhemmat M B Khamesee andF Golnaraghi ldquoA hybrid electromagnetic shock absorber foractive vehicle suspension systemsrdquo Vehicle System Dynamicsvol 49 no 1-2 pp 311ndash332 2011
[12] S Tang L Chen R Wang X Sun and D Shi ldquoResearch onoptimal control of active suspension based on damping multi-modal switchingrdquo Journal of Guangxi University (NaturalScience) vol 39 no 2 pp 300ndash307 2014
[13] R Wang X Ma R Ding X Meng and L Chen ldquoResearchof multi-mode switching control system for hybrid sus-pension based on model referencerdquo Transactions of theChinese Society for Agricultural Machinery vol 48 no 7pp 353ndash360 2017
[14] R Wang Y Qian R Ding X Meng and J Xie ldquoDesign andtests for damping-stiffness of a hybrid electromagnetic sus-pension based on LQGrdquo Journal of Vibration and Shockvol 37 no 3 pp 61ndash65 2017
[15] B Vanavil K K Chaitanya and A S Rao ldquoImproved PIDcontroller design for unstable time delay processes based ondirect synthesis method and maximum sensitivityrdquo Taylorand Francis vol 46 no 8 pp 1349ndash1366 2015
[16] I Mihai and F Andronic ldquoBehavior of a semi-active sus-pension system versus a passive suspension system on anuneven road surfacerdquo Mechanics vol 20 no 1 pp 64ndash692014
[17] D Ngoduy ldquoLinear stability of a generalized multi-anticipative car following model with time delaysrdquo Com-munications in Nonlinear Science and Numerical Simulationvol 22 no 1 pp 420ndash426 2015
[18] H Zhang E Wang F Min R Subash and C Su ldquoSkyhook-based semi-active control of full-vehicle suspension with
magneto-rheological dampersrdquo Chinese Journal of Mechan-ical Engineering vol 26 no 3 pp 498ndash505 2013
[19] F Kou ldquoAn experimental study on the dynamic character-istics of vehicle semi-active seat suspension with magneto-rheological damperrdquo Automotive Engineering vol 37 no 11pp 1346ndash1352 2015
[20] Z Feng S Chen and Y Liang ldquoAn experimental study on thedynamic characteristics of a megneto-rheological semi-activesuspensionrdquoAutomotive Engineering vol 35 no 1 pp 72ndash772013
[21] Z Li and L Xu A New Type of Magnetorheological Damperand Semi-Active coNtrol Design-eory Science Press BeijingChina 2012
[22] B L J Gysen J J H Paulides J L G Janssen andE A Lomonova ldquoActive electromagnetic suspension systemfor improved vehicle dynamicsrdquo IEEE Transactions on Ve-hicular Technology vol 59 no 3 pp 1156ndash1163 2010
[23] Q N Wang S S Liu W H Wang and H Wei ldquoStructuredesign and parameter matching of ball-screw regenerativedamperrdquo Journal of Jilin University (Engineering and Tech-nology Edition) vol 42 no 5 pp 1100ndash1106 2012
[24] S Chen R He and S Lu ldquoEvaluating system of reclaimingenergy suspension comprehensive performancerdquo Trans-actions of the Chinese Society for Agricultural Machineryvol 37 no 7 pp 14ndash18 2006
[25] Y Fan Control Research on Vehicle Suspension with Electro-Hydrostatic Actuator Xirsquoan University of Science and Tech-nology Xirsquoan China 2017
[26] Z Yu Automobile -eory China Machine Press BeijingChina 2009
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Navigation and Observation
International Journal of
Hindawi
wwwhindawicom Volume 2018
Advances in
Multimedia
Submit your manuscripts atwwwhindawicom
the sky-hook coecient csky is constant However at thistime the ball screw actuator as a power feeding devicegenerates the electromagnetic damping force Fs and acts onthe suspension so that the actual semiactive control force ofthe suspension is dierent from the ideal semiactive controlforce minus csky middot _x2 In this paper the semiactive control modelof the hybrid suspension is established and the changeeects of the dierent output forces on the vehicle ridingcomfort and handling stability are analyzed by MATLABSimulink software e simulation speed is 70 kmh thesimulation time is 5 s and the value of R is 075 Ω edamping comparison of the hybrid suspension in semiactivecontrol is shown in Figure 14
Figure 14 shows that compared with the ideal semi-active control force the actual semiactive control force ofthe hybrid suspension shyuctuates violently and the abso-lute value of the actual semiactive control force is greaterthan the absolute value of the ideal semiactive control force|csky middot _x2| at certain times And a drastic change in theactual semiactive control force makes the suspension notreach ideal semiactive control eect Using electromag-netic damping force feedback adjustment to reduce thedierence between the ideal semiactive control forceminus csky middot _x2 and the actual semiactive control force themethod is as follows
When |csky middot _x2|gt |Fs| the semiactive control force of thehybrid suspension is provided by both the MR damper andthe ball screw actuator and at this point the controllerinputs a controllable current Ik to the MR damper so thatthe Fk output by the MR damper is minus csky middot _x2 minusFs Andwhen |csky middot _x2|le |Fs| the semiactive control force of thehybrid suspension is the Fs which is output by the ball screwactuator and at this point there is no controllable current Ikinput to theMR damper and the function of theMR damperis equivalent to a traditional shock absorber erefore thesemiactive control of the hybrid suspension does not havethe dead zone of traditional electromagnetic semiactivesuspension which helps to improve the semiactive controleect of the hybrid suspension
When there is feedback adjustment the semiactivecontrol force of the hybrid suspension is
Fb Fs Fs
∣∣∣∣∣∣∣∣ge csky middot _x2∣∣∣∣∣
∣∣∣∣∣
minus csky middot _x2 Fs∣∣∣∣∣∣∣∣lt csky middot _x2∣∣∣∣∣
∣∣∣∣∣
(32)
When there is feedback adjustment the Fk output by theMR damper is
Fk 0 Fs
∣∣∣∣∣∣∣∣ge csky middot _x2∣∣∣∣∣
∣∣∣∣∣
minus csky middot _x2 minusFs Fs∣∣∣∣∣∣∣∣lt csky middot _x2∣∣∣∣∣
∣∣∣∣∣
(33)
From equations (31)sim(33) when the electromagneticdamping force feedback adjustment is used the |Fk| outputby theMR damper decreases and when |Fk| decreases it canbe known from equations (8) and (9) that the energyconsumption of the MR damper decreases with it
e comparison of the semiactive control force of thehybrid suspension with or without the electromagneticdamping force feedback adjustment is shown in Figure 15
From Figure 15 the RMS of the ideal semiactive controlforce of the hybrid suspension is 3276 N and when there isno electromagnetic damping force feedback adjustmentthe RMS of the actual semiactive control force of thesuspension is 4041 N and the dierence between the actualsemiactive force of the suspension and the ideal semiactiveforce is 2335 When there is electromagnetic dampingforce feedback adjustment the RMS of the actual semi-active control force of the suspension is 3593 N and thedierence between the actual semiactive force of the sus-pension and the ideal semiactive force is 968 ereforewhen there is electromagnetic damping force feedbackadjustment the actual semiactive control force of thesuspension has a smaller shyuctuation amplitude whichhelps to improve the semiactive control eect of the hybridsuspension
e dynamic responses of the hybrid suspension with orwithout electromagnetic damping force feedback adjust-ment are shown in Figure 16 Among them the damper ofthe passive suspension is the original damper of the vehicleand its damping value is 1600Nmiddotsm
Table 2 shows the response RMS values of the hybridsuspension in semiactive control
From Table 2 compared with the passive suspensionwhen there is electromagnetic damping force feedbackadjustment aw SWSrms and DTLrms of the hybrid sus-pension are reduced by 1698 432 and 1068 re-spectively and compared with the nonfeedback semiactivecontrol when the feedback semiactive control is performedaw SWSrms and DTLrms of the hybrid suspension are re-duced by 252 863 and 671 respectively
From equations (9) and (29) the total system energy ofthe hybrid suspension in semiactive control is
W1 Wk minusWb (34)
whereW1 is the total system energy of the hybrid suspensionin semiactive control
0 1 2 3 4 5ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Hyb
rid su
spen
sion
forc
e (N
)
072 076 080
300
600
Ideal semiactive forceActual semiactive forceElectromagnetic damping force
Figure 14 Damping comparison of the hybrid suspension insemiactive control
8 Shock and Vibration
0 1 2 3 4 5ndash10
ndash5
0
5
10
15
Time (s)
Spru
ng m
ass a
ccel
erat
ion
(ms
2 )
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(a)
0 1 2 3 4 5ndash004
ndash002
000
002
004
006
008
Time (s)
Susp
ensio
n w
orki
ng sp
ace (
m)
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(b)
0 1 2 3 4 5ndash3000
ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Dyn
amic
tire
load
(N)
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(c)
Figure 16 e dynamic responses of the hybrid suspension in semiactive control (a) e response curves of sprung mass acceleration (b)e response curves of suspension working space (c) e response curves of dynamic tire load
0 1 2 3 4 5ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Hyb
rid su
spen
sion
forc
e (N
)
072 076 080
300
600
Ideal semiactive forceWithout feedback semiactive forceWith feedback semiactive force
Figure 15 Semiactive force of the hybrid suspension
Shock and Vibration 9
From equation (33) when the hybrid suspension sem-iactive controlled the curves of the total system energychange over time are shown in Figure 17
From Figure 17 when there is the nonfeedback semiactivecontrol the total system energy of the hybrid suspension is60 J And when there is the feedback semiactive control thetotal system energy of the hybrid suspension is 307 J
43 -e Design of Multimode Coordination ControllerWhen the hybrid suspension is actively controlled thesuspension has good vibration isolation performance buthigh energy consumption And when the hybrid suspensionis semiactively controlled the suspension has good eco-nomic performance but the control has limitations Con-sidering that the ball screw actuator can realize active controlof the hybrid suspension in any suspension state a multi-mode coordinated control strategy of the hybrid suspensionis designed
minus csky middot _x2 Fs1113868111386811138681113868
1113868111386811138681113868gt csky middot _x2
11138681113868111386811138681113868
11138681113868111386811138681113868
⎧⎪⎨
⎪⎩
cs c0
⎧⎪⎪⎪⎨
⎪⎪⎪⎩
_x2 minus _x1 1113857 _x2 gt 0
F Fz minus csky middot _x2
cs c0 + ck1113896 _x2 minus _x1 1113857 _x2 le 0
⎧⎪⎪⎪⎪⎪⎪⎪⎪⎪⎨
⎪⎪⎪⎪⎪⎪⎪⎪⎪⎩
(35)
0e frame diagram of the multimode coordinated controlstrategy of the hybrid suspension is shown in Figure 18
From equations (9) (12) (26) and (29) the total systemenergy of the hybrid suspension in multimode coordinatedcontrol is
W2 Wk minusWb minusWZ minusWc (36)
whereW2 is the total system energy of the hybrid suspensionin multimode coordinated control
A hybrid active suspension simulation model is estab-lished by using MATLABSimulink software From Fig-ure 11 when the hybrid suspension is actively controlled theenergy consumption of the MR damper is mainly affected bythe vehicle speed In order to verify the vibration isolationperformance and energy consumption performance of thehybrid active suspension in the cyclic driving conditionsthis article simulates vehicle urban and suburban conditionswhich is based on GBT 19233-2003 ldquoLight Vehicle FuelConsumption Test Methodrdquo among them the urban con-ditions include four cycle units each cycle time is 195 s thesuburban conditions include one cycle unit and the cycletime is 400 s [26] And the schematic diagram of the urbanconditions unit is shown in Figure 19
In order to simulate the vehicle acceleration signal avehicle speed variation model in different cycle units isestablished by using the signal builder function module inSimulink software On this basis the vehicle accelerationmodel in different cycle units is obtained and the vehicleacceleration model can be used as the switching controlmodel for the variable damping when the hybrid suspensionis actively controlled 0e shock absorbers of active sus-pension and passive suspension adopt the original damper ofvehicle and the damping value c1 is 1600 Nmiddotsm Howeverbecause of the long simulation time the dynamic responsecurves of the hybrid active suspension in urban and sub-urban conditions cannot be displayed Figure 20 shows thedynamic response curve of the hybrid active suspension inthe 145ndash175 s in the urban circulation unit among themvehicle in the 145ndash155 s is in a constant speed vehicle in the155ndash163 s is in the deceleration state and vehicle in the163ndash175 s is in other constant speed
0e dynamic response RMS values of the hybrid sus-pension in urban and suburban circulation units are shownin Tables 3 and 4 respectively
From Figure 20 and Tables 3 and 4 the dynamic re-sponses are good when the hybrid suspension is multimodecoordinated control When in the urban circulation unitcompared with the passive suspension aw SWSrms andDTLrms of the hybrid suspension are reduced by 3943225 and 2081 respectively Compared with the activesuspension aw and DTLrms of the hybrid suspension arereduced by 1322 and 840 respectively but SWSrms isincreased by 805 When in the suburban circulation unitcompared with the passive suspension aw SWSrms andDTLrms of the hybrid suspension are reduced by 3916314 and 1955 respectively Compared with the activesuspension aw and DTLrms of the hybrid suspension arereduced by 1199 and 732 respectively but SWSrms isincreased by 845
0e system energy of the hybrid suspension in urban andsuburban circulation units are shown in Figures 21 and 22respectively
Table 2 0e response root mean square values of the hybridsuspension in semiactive control
Without feedback semiactive controlWith feedback semiactive control
Figure 17 0e system energy of the hybrid suspension in semi-active control
10 Shock and Vibration
en the system energy values of the hybrid suspensionin urban and suburban circulation units are shown inTable 5
From Table 5 when in the urban circulation unit thesystem energy of the active suspension and hybrid sus-pension is minus2165 J and 38 J respectively And when in thesuburban circulation unit the system energy of the activesuspension and hybrid suspension is minus15071 J and minus122 Jrespectively e entire operation cycle includes 4 urbancycle units and 1 suburban cycle unit so the pure energy ofthe active suspension system during the entire operationcycle is minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid suspensionsystem basically realizes energy self-powered in theory
5 Test and Analysis
In order to verify the damping eect of the hybrid sus-pension system in active mode and semiactive mode ahybrid suspension vibration test system is designed as
Suspension controllerSpeed of sprung mass x2
Speed of unsprung mass x1Vehicle acceleration v
Vehicle suspension
Ball screw actuator
Output Fk
Active control force Fz
Ball screw actuator
Output Iz
Output Ik
Electromagnetic damping force Fs
Ideal semiactive control force FL
Output semiactive controlforce Fb = Fs
Output semiactive controlforce Fb = FL = Fk + Fs
x2 ndash x1
(x2 ndash x1) x2 gt 0
(x2 ndash x1) x2 le 0
Output Ik
MR damper
Adjust the duty cycle in real timeaccording to the relationshipbetween active output force
and duty cycle
No controllablecurrent output
MR damper
v ne 0
According to vand ∆v
v = 0
MR damper
Adjustable damping cs = c0 + ck
ndashcsky middot x2
ndashcsky middot x2
|Fs| ge |FL|
|Fs| lt |FL|
Figure 18 e frame diagram of the multimode coordinated control strategy of the hybrid suspension
Vehicle shifting
00
15
10
30
30
20
45
40
50
60
60
75 90 105 120Time (s)
Veh
icle
spee
d (k
mh
)
135 150 165 180 195
Figure 19 e schematic diagram of the urban conditions unit
Shock and Vibration 11
shown in Figure 23 During the test the MR damper has nocontrollable current input in the active control mode of thehybrid suspension and its damping value is always 800Nmiddotsm
e passive suspension semiactive suspension and activesuspension all adopt the original damper of the vehicle andthe damping value c1 is 1600Nmiddotsm
Figure 20e dynamic responses of the hybrid suspension (a)e response curves of sprungmass acceleration (b)e response curves ofsuspension working space (c) e response curves of dynamic tire load
12 Shock and Vibration
Because of the limitation of test conditions only thesprung mass acceleration dynamic response of the hybridsuspension is measured in this test e dynamic response ofthe sprung mass acceleration of the hybrid suspension inactive mode under random road is shown in Figure 24 epower spectrum of the sprung mass acceleration of thehybrid suspension in active mode is shown in Figure 25
Table 6 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in active mode
From Table 6 compared with passive suspension aw ofthe hybrid suspension in active mode is reduced by 3945and compared with active suspension aw of the hybridsuspension in active mode is reduced by 1432 And fromFigure 25 when the hybrid suspension is actively controlledcompared with passive suspension the vibration isolationperformance of the suspension in the low frequency andlow-frequency resonance areas is similar to that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withactive suspension the vibration isolation performance of thesuspension in the low frequency high frequency and res-onance regions is slightly better
e dynamic response of the sprung mass acceleration ofthe hybrid suspension in semiactive mode under randomroad is shown in Figure 26 e power spectrum of thesprung mass acceleration of the hybrid suspension insemiactive mode is shown in Figure 27
Table 7 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in semiactive mode
From Table 7 compared with passive suspension aw ofthe hybrid suspension in semiactive mode is reduced by1642 Compared with semiactive suspension aw of thehybrid suspension in active mode is reduced by 307 Andfrom Figure 27 when the hybrid suspension is semiactivelycontrolled compared with passive suspension the vibration
Table 3 e dynamic response root mean square values of thehybrid suspension in urban circulation unit
Figure 21 e system energy of the hybrid suspension in urbancirculation units
0 100 200 300 400ndash20000
ndash15000
ndash10000
ndash5000
0
5000
Time (s)
Susp
ensio
n sy
stem
ener
gy (J
)
Active suspensionHybrid suspension
Figure 22e system energy of the hybrid suspension in suburbancirculation units
Table 5 e system energy values of the hybrid suspension (J)
Cycle unit Active suspension Hybrid suspensionUrban minus2165 38Suburban minus15071 minus122
Figure 23 e test system of the hybrid active suspension
Shock and Vibration 13
isolation performance of the suspension in the low frequencyand low-frequency resonance areas is worse than that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withsemiactive suspension the vibration isolation performance ofthe suspension in the low frequency high frequency andresonance regions is slightly better
e energy consumption power and energy-regenerativepower of the hybrid suspension are shown in Figure 28
From Figure 28 in the active mode the average powerconsumption of the ball screw actuator is 2561W In thesemiactive mode the average regenerative power of the ballscrew actuator is 2696W and the average of theMR damperenergy consumption power is 178W erefore the pureaverage power consumption of the hybrid suspension isminus043W and the test results are basically consistent with thesimulation
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spr
ung
mas
s acc
eler
atio
n (m
s2 )
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 24 e dynamic response of the sprung mass accelerationof the hybrid suspension in active mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 25 e power spectrum of the sprung mass acceleration ofthe hybrid suspension in active mode
Table 6 e sprung mass acceleration RMS of the hybrid sus-pension in active mode
Indicators Passivesuspension
Activesuspension
Hybridsuspension
aw (ms2) 13378 09423 08074
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spru
ng m
ass a
ccel
erat
ion
(ms
2 )
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 26 e dynamic responses of the sprung mass accelerationof the hybrid suspension in semiactive mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 27e power spectrums of the sprungmass acceleration ofthe hybrid suspension in semiactive mode
Table 7 e sprung mass acceleration RMS of the hybrid sus-pension in semiactive mode
Indicators Passivesuspension
Semiactivesuspension
Hybridsuspension
aw (ms2) 13378 11536 11182
14 Shock and Vibration
6 Conclusion
(1) A ball screw actuator andMR damper are introducedinto vehicle suspension system and a new kind ofhybrid active suspension structure is put forwarde ball screw actuator is prototyped and thefunctional relationship between the back-EMF co-ecient the electromagnetic torque coecient of themotor and the suspension vibration speed is ob-tained by test analyses And the active output me-chanical properties of the ball screw actuator aretested and the results show that the actuator hasgood active output force characteristics
(2) e inshyuences of the variable damping value of thesuspension system on the riding comfort handlingand stability and energy consumption characteristicsof the hybrid suspension in the active control modeare analyzeden the optimal damping values of thehybrid suspension at dierent vehicle speeds aredesigned e eects of electromagnetic dampingforce on the actual semiactive force and the systemenergy of the suspension in the semiactive controlmode are analyzed and then the hybrid suspensionwith semiactive mode which has electromagneticdamping force feedback adjustment is designed Onthis basis a multimode coordinated control strategyfor the hybrid suspension is designed
(3) e damping performance and energy consumptioncharacteristics of the hybrid suspension under cyclicdriving condition are simulated by MATLABSimulink software and the results show that whenin the urban circulation unit compared with activesuspension aw and DTLrms of the hybrid suspensionare reduced by 1322 and 840 respectively Andwhen in the suburban circulation unit comparedwith the active suspension aw and DTLrms of thehybrid suspension are reduced by 1199 and 732respectively e pure energy of the active suspen-sion system during the entire operation cycle is
minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid sus-pension system basically realizes energy self-poweredin theory
(4) e eectiveness verication test of the hybrid sus-pension in active mode and semiactive mode controlis carried out and the results show that when thehybrid suspension is actively controlled comparedwith active suspension aw of the hybrid suspension inactive mode is reduced by 1432 When the hybridsuspension is semiactively controlled compared withsemiactive suspension aw of the hybrid suspension inactive mode is reduced by 307 e test and sim-ulation results are basically consistent and the testveries the correctness of the simulation
Data Availability
e data used to support the ndings of this study are in-cluded within the supplementary information les And thedata used to support the ndings of this study are availablefrom the corresponding author upon request
Conflicts of Interest
e authors declare that there are no conshyicts of interestregarding the publication of this paper
Acknowledgments
is work was supported by the National Natural ScienceFoundation of China (Grant no 51775426) Service LocalSpecial Program Support Project of Shaanxi ProvincialEducation Department (Grant no 17JF017) and XirsquoanScience and Technology Program Funding Project (Grantno 2017079CGRC042-XAKD007)
Supplementary Materials
e supplementary materials are this articlersquos experimentaldata including (1) MR damperrsquos characteristic test data (2)the motor counter electromotive forcersquos peak test data (3)the ball screw actuatorrsquos active output force test data (4) thehybrid suspensionrsquos sprung mass acceleration test data inactive mode (5) the hybrid suspensionrsquos sprung mass ac-celeration test data in semiactive mode and (6) the hybridsuspensionrsquos energy consumption power and energy re-generation power test data (Supplementary Materials)
References
[1] M Montazeri-Gh and O Kavianipour ldquoInvestigation of theactive electromagnetic suspension system considering hybridcontrol strategyrdquo Proceedings of the Institution of MechanicalEngineers Part C Journal of Mechanical Engineering Sciencevol 228 no 10 pp 1658ndash1669 2013
[2] H E Tseng and D Hrovat ldquoState of the art survey active andsemi-active suspension controlrdquo Vehicle System Dynamicsvol 53 no 7 pp 1034ndash1062 2015
0 1 2 3 4 50
20
40
60
80
Time (s)
Pow
er (W
)
Energy consumption power in active modeEnergy regenerative power in semiactive modeEnergy consumption power in semiactive mode
Figure 28 e energy consumption power and energy-re-generative power of the hybrid suspension
Shock and Vibration 15
[3] F Kou J Du Z Wang D Li and J Xu ldquoNonlinear modelingand coordinate optimization of a semi-active energy re-generative suspension with an electro-hydraulic actuatorrdquoChina Mechanical Engineering vol 28 no 14 pp 1701ndash17072017
[4] S H Zareh A Sarrafan A A A Khayyat and A ZabihollahldquoIntelligent semi-active vibration control of eleven degrees offreedom suspension system using magnetorheologicaldampersrdquo Journal of Mechanical Science and Technologyvol 26 no 3 pp 323ndash334 2012
[5] S A Chen X Li L J Zhao Y X Wang and Y B KimldquoDevelopment of a control method for an electromagneticsemi-active suspension reclaiming energy with varying chargevoltage in stepsrdquo International Journal of Automotive Tech-nology vol 16 no 5 pp 765ndash773 2015
[6] K Nakano Y Suda and S Nakadai ldquoSelf-powered activevibration control using a single electric actuatorrdquo Journal ofSound and Vibration vol 260 no 2 pp 213ndash235 2003
[7] K Nakano ldquoCombined type self-powered active vibrationcontrol of truck cabinsrdquo Vehicle System Dynamics vol 41no 6 pp 449ndash473 2004
[8] K Huang Y C Zhang F Yu and Y H Gu ldquoCoordinateoptimization for synthetical performance of electrical energy-regenerative active suspensionrdquo Journal of Shanghai JiaotongUniversity vol 43 no 2 pp 226ndash230 2009
[9] K Huang F Yu and Y C Zhang ldquoActive control of energy-regenerative electromagnetic suspension based on energy flowanalysisrdquo Journal of Shanghai Jiaotong University vol 45no 67 pp 1068ndash1073 2011
[10] D S Huang J Q Zhang Y L Liu L Yi and X Y WangldquoPerformance of a novel energy-regenerative active suspen-sion systemrdquo Journal of Chongqing University (English Edi-tion) vol 14 no 3 pp 109ndash118 2015
[11] B Ebrahimi H Bolandhemmat M B Khamesee andF Golnaraghi ldquoA hybrid electromagnetic shock absorber foractive vehicle suspension systemsrdquo Vehicle System Dynamicsvol 49 no 1-2 pp 311ndash332 2011
[12] S Tang L Chen R Wang X Sun and D Shi ldquoResearch onoptimal control of active suspension based on damping multi-modal switchingrdquo Journal of Guangxi University (NaturalScience) vol 39 no 2 pp 300ndash307 2014
[13] R Wang X Ma R Ding X Meng and L Chen ldquoResearchof multi-mode switching control system for hybrid sus-pension based on model referencerdquo Transactions of theChinese Society for Agricultural Machinery vol 48 no 7pp 353ndash360 2017
[14] R Wang Y Qian R Ding X Meng and J Xie ldquoDesign andtests for damping-stiffness of a hybrid electromagnetic sus-pension based on LQGrdquo Journal of Vibration and Shockvol 37 no 3 pp 61ndash65 2017
[15] B Vanavil K K Chaitanya and A S Rao ldquoImproved PIDcontroller design for unstable time delay processes based ondirect synthesis method and maximum sensitivityrdquo Taylorand Francis vol 46 no 8 pp 1349ndash1366 2015
[16] I Mihai and F Andronic ldquoBehavior of a semi-active sus-pension system versus a passive suspension system on anuneven road surfacerdquo Mechanics vol 20 no 1 pp 64ndash692014
[17] D Ngoduy ldquoLinear stability of a generalized multi-anticipative car following model with time delaysrdquo Com-munications in Nonlinear Science and Numerical Simulationvol 22 no 1 pp 420ndash426 2015
[18] H Zhang E Wang F Min R Subash and C Su ldquoSkyhook-based semi-active control of full-vehicle suspension with
magneto-rheological dampersrdquo Chinese Journal of Mechan-ical Engineering vol 26 no 3 pp 498ndash505 2013
[19] F Kou ldquoAn experimental study on the dynamic character-istics of vehicle semi-active seat suspension with magneto-rheological damperrdquo Automotive Engineering vol 37 no 11pp 1346ndash1352 2015
[20] Z Feng S Chen and Y Liang ldquoAn experimental study on thedynamic characteristics of a megneto-rheological semi-activesuspensionrdquoAutomotive Engineering vol 35 no 1 pp 72ndash772013
[21] Z Li and L Xu A New Type of Magnetorheological Damperand Semi-Active coNtrol Design-eory Science Press BeijingChina 2012
[22] B L J Gysen J J H Paulides J L G Janssen andE A Lomonova ldquoActive electromagnetic suspension systemfor improved vehicle dynamicsrdquo IEEE Transactions on Ve-hicular Technology vol 59 no 3 pp 1156ndash1163 2010
[23] Q N Wang S S Liu W H Wang and H Wei ldquoStructuredesign and parameter matching of ball-screw regenerativedamperrdquo Journal of Jilin University (Engineering and Tech-nology Edition) vol 42 no 5 pp 1100ndash1106 2012
[24] S Chen R He and S Lu ldquoEvaluating system of reclaimingenergy suspension comprehensive performancerdquo Trans-actions of the Chinese Society for Agricultural Machineryvol 37 no 7 pp 14ndash18 2006
[25] Y Fan Control Research on Vehicle Suspension with Electro-Hydrostatic Actuator Xirsquoan University of Science and Tech-nology Xirsquoan China 2017
[26] Z Yu Automobile -eory China Machine Press BeijingChina 2009
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Navigation and Observation
International Journal of
Hindawi
wwwhindawicom Volume 2018
Advances in
Multimedia
Submit your manuscripts atwwwhindawicom
0 1 2 3 4 5ndash10
ndash5
0
5
10
15
Time (s)
Spru
ng m
ass a
ccel
erat
ion
(ms
2 )
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(a)
0 1 2 3 4 5ndash004
ndash002
000
002
004
006
008
Time (s)
Susp
ensio
n w
orki
ng sp
ace (
m)
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(b)
0 1 2 3 4 5ndash3000
ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Dyn
amic
tire
load
(N)
Passive suspensionWithout feedback semiactive controlWith feedback semiactive control
(c)
Figure 16 e dynamic responses of the hybrid suspension in semiactive control (a) e response curves of sprung mass acceleration (b)e response curves of suspension working space (c) e response curves of dynamic tire load
0 1 2 3 4 5ndash2000
ndash1000
0
1000
2000
3000
Time (s)
Hyb
rid su
spen
sion
forc
e (N
)
072 076 080
300
600
Ideal semiactive forceWithout feedback semiactive forceWith feedback semiactive force
Figure 15 Semiactive force of the hybrid suspension
Shock and Vibration 9
From equation (33) when the hybrid suspension sem-iactive controlled the curves of the total system energychange over time are shown in Figure 17
From Figure 17 when there is the nonfeedback semiactivecontrol the total system energy of the hybrid suspension is60 J And when there is the feedback semiactive control thetotal system energy of the hybrid suspension is 307 J
43 -e Design of Multimode Coordination ControllerWhen the hybrid suspension is actively controlled thesuspension has good vibration isolation performance buthigh energy consumption And when the hybrid suspensionis semiactively controlled the suspension has good eco-nomic performance but the control has limitations Con-sidering that the ball screw actuator can realize active controlof the hybrid suspension in any suspension state a multi-mode coordinated control strategy of the hybrid suspensionis designed
minus csky middot _x2 Fs1113868111386811138681113868
1113868111386811138681113868gt csky middot _x2
11138681113868111386811138681113868
11138681113868111386811138681113868
⎧⎪⎨
⎪⎩
cs c0
⎧⎪⎪⎪⎨
⎪⎪⎪⎩
_x2 minus _x1 1113857 _x2 gt 0
F Fz minus csky middot _x2
cs c0 + ck1113896 _x2 minus _x1 1113857 _x2 le 0
⎧⎪⎪⎪⎪⎪⎪⎪⎪⎪⎨
⎪⎪⎪⎪⎪⎪⎪⎪⎪⎩
(35)
0e frame diagram of the multimode coordinated controlstrategy of the hybrid suspension is shown in Figure 18
From equations (9) (12) (26) and (29) the total systemenergy of the hybrid suspension in multimode coordinatedcontrol is
W2 Wk minusWb minusWZ minusWc (36)
whereW2 is the total system energy of the hybrid suspensionin multimode coordinated control
A hybrid active suspension simulation model is estab-lished by using MATLABSimulink software From Fig-ure 11 when the hybrid suspension is actively controlled theenergy consumption of the MR damper is mainly affected bythe vehicle speed In order to verify the vibration isolationperformance and energy consumption performance of thehybrid active suspension in the cyclic driving conditionsthis article simulates vehicle urban and suburban conditionswhich is based on GBT 19233-2003 ldquoLight Vehicle FuelConsumption Test Methodrdquo among them the urban con-ditions include four cycle units each cycle time is 195 s thesuburban conditions include one cycle unit and the cycletime is 400 s [26] And the schematic diagram of the urbanconditions unit is shown in Figure 19
In order to simulate the vehicle acceleration signal avehicle speed variation model in different cycle units isestablished by using the signal builder function module inSimulink software On this basis the vehicle accelerationmodel in different cycle units is obtained and the vehicleacceleration model can be used as the switching controlmodel for the variable damping when the hybrid suspensionis actively controlled 0e shock absorbers of active sus-pension and passive suspension adopt the original damper ofvehicle and the damping value c1 is 1600 Nmiddotsm Howeverbecause of the long simulation time the dynamic responsecurves of the hybrid active suspension in urban and sub-urban conditions cannot be displayed Figure 20 shows thedynamic response curve of the hybrid active suspension inthe 145ndash175 s in the urban circulation unit among themvehicle in the 145ndash155 s is in a constant speed vehicle in the155ndash163 s is in the deceleration state and vehicle in the163ndash175 s is in other constant speed
0e dynamic response RMS values of the hybrid sus-pension in urban and suburban circulation units are shownin Tables 3 and 4 respectively
From Figure 20 and Tables 3 and 4 the dynamic re-sponses are good when the hybrid suspension is multimodecoordinated control When in the urban circulation unitcompared with the passive suspension aw SWSrms andDTLrms of the hybrid suspension are reduced by 3943225 and 2081 respectively Compared with the activesuspension aw and DTLrms of the hybrid suspension arereduced by 1322 and 840 respectively but SWSrms isincreased by 805 When in the suburban circulation unitcompared with the passive suspension aw SWSrms andDTLrms of the hybrid suspension are reduced by 3916314 and 1955 respectively Compared with the activesuspension aw and DTLrms of the hybrid suspension arereduced by 1199 and 732 respectively but SWSrms isincreased by 845
0e system energy of the hybrid suspension in urban andsuburban circulation units are shown in Figures 21 and 22respectively
Table 2 0e response root mean square values of the hybridsuspension in semiactive control
Without feedback semiactive controlWith feedback semiactive control
Figure 17 0e system energy of the hybrid suspension in semi-active control
10 Shock and Vibration
en the system energy values of the hybrid suspensionin urban and suburban circulation units are shown inTable 5
From Table 5 when in the urban circulation unit thesystem energy of the active suspension and hybrid sus-pension is minus2165 J and 38 J respectively And when in thesuburban circulation unit the system energy of the activesuspension and hybrid suspension is minus15071 J and minus122 Jrespectively e entire operation cycle includes 4 urbancycle units and 1 suburban cycle unit so the pure energy ofthe active suspension system during the entire operationcycle is minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid suspensionsystem basically realizes energy self-powered in theory
5 Test and Analysis
In order to verify the damping eect of the hybrid sus-pension system in active mode and semiactive mode ahybrid suspension vibration test system is designed as
Suspension controllerSpeed of sprung mass x2
Speed of unsprung mass x1Vehicle acceleration v
Vehicle suspension
Ball screw actuator
Output Fk
Active control force Fz
Ball screw actuator
Output Iz
Output Ik
Electromagnetic damping force Fs
Ideal semiactive control force FL
Output semiactive controlforce Fb = Fs
Output semiactive controlforce Fb = FL = Fk + Fs
x2 ndash x1
(x2 ndash x1) x2 gt 0
(x2 ndash x1) x2 le 0
Output Ik
MR damper
Adjust the duty cycle in real timeaccording to the relationshipbetween active output force
and duty cycle
No controllablecurrent output
MR damper
v ne 0
According to vand ∆v
v = 0
MR damper
Adjustable damping cs = c0 + ck
ndashcsky middot x2
ndashcsky middot x2
|Fs| ge |FL|
|Fs| lt |FL|
Figure 18 e frame diagram of the multimode coordinated control strategy of the hybrid suspension
Vehicle shifting
00
15
10
30
30
20
45
40
50
60
60
75 90 105 120Time (s)
Veh
icle
spee
d (k
mh
)
135 150 165 180 195
Figure 19 e schematic diagram of the urban conditions unit
Shock and Vibration 11
shown in Figure 23 During the test the MR damper has nocontrollable current input in the active control mode of thehybrid suspension and its damping value is always 800Nmiddotsm
e passive suspension semiactive suspension and activesuspension all adopt the original damper of the vehicle andthe damping value c1 is 1600Nmiddotsm
Figure 20e dynamic responses of the hybrid suspension (a)e response curves of sprungmass acceleration (b)e response curves ofsuspension working space (c) e response curves of dynamic tire load
12 Shock and Vibration
Because of the limitation of test conditions only thesprung mass acceleration dynamic response of the hybridsuspension is measured in this test e dynamic response ofthe sprung mass acceleration of the hybrid suspension inactive mode under random road is shown in Figure 24 epower spectrum of the sprung mass acceleration of thehybrid suspension in active mode is shown in Figure 25
Table 6 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in active mode
From Table 6 compared with passive suspension aw ofthe hybrid suspension in active mode is reduced by 3945and compared with active suspension aw of the hybridsuspension in active mode is reduced by 1432 And fromFigure 25 when the hybrid suspension is actively controlledcompared with passive suspension the vibration isolationperformance of the suspension in the low frequency andlow-frequency resonance areas is similar to that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withactive suspension the vibration isolation performance of thesuspension in the low frequency high frequency and res-onance regions is slightly better
e dynamic response of the sprung mass acceleration ofthe hybrid suspension in semiactive mode under randomroad is shown in Figure 26 e power spectrum of thesprung mass acceleration of the hybrid suspension insemiactive mode is shown in Figure 27
Table 7 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in semiactive mode
From Table 7 compared with passive suspension aw ofthe hybrid suspension in semiactive mode is reduced by1642 Compared with semiactive suspension aw of thehybrid suspension in active mode is reduced by 307 Andfrom Figure 27 when the hybrid suspension is semiactivelycontrolled compared with passive suspension the vibration
Table 3 e dynamic response root mean square values of thehybrid suspension in urban circulation unit
Figure 21 e system energy of the hybrid suspension in urbancirculation units
0 100 200 300 400ndash20000
ndash15000
ndash10000
ndash5000
0
5000
Time (s)
Susp
ensio
n sy
stem
ener
gy (J
)
Active suspensionHybrid suspension
Figure 22e system energy of the hybrid suspension in suburbancirculation units
Table 5 e system energy values of the hybrid suspension (J)
Cycle unit Active suspension Hybrid suspensionUrban minus2165 38Suburban minus15071 minus122
Figure 23 e test system of the hybrid active suspension
Shock and Vibration 13
isolation performance of the suspension in the low frequencyand low-frequency resonance areas is worse than that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withsemiactive suspension the vibration isolation performance ofthe suspension in the low frequency high frequency andresonance regions is slightly better
e energy consumption power and energy-regenerativepower of the hybrid suspension are shown in Figure 28
From Figure 28 in the active mode the average powerconsumption of the ball screw actuator is 2561W In thesemiactive mode the average regenerative power of the ballscrew actuator is 2696W and the average of theMR damperenergy consumption power is 178W erefore the pureaverage power consumption of the hybrid suspension isminus043W and the test results are basically consistent with thesimulation
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spr
ung
mas
s acc
eler
atio
n (m
s2 )
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 24 e dynamic response of the sprung mass accelerationof the hybrid suspension in active mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 25 e power spectrum of the sprung mass acceleration ofthe hybrid suspension in active mode
Table 6 e sprung mass acceleration RMS of the hybrid sus-pension in active mode
Indicators Passivesuspension
Activesuspension
Hybridsuspension
aw (ms2) 13378 09423 08074
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spru
ng m
ass a
ccel
erat
ion
(ms
2 )
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 26 e dynamic responses of the sprung mass accelerationof the hybrid suspension in semiactive mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 27e power spectrums of the sprungmass acceleration ofthe hybrid suspension in semiactive mode
Table 7 e sprung mass acceleration RMS of the hybrid sus-pension in semiactive mode
Indicators Passivesuspension
Semiactivesuspension
Hybridsuspension
aw (ms2) 13378 11536 11182
14 Shock and Vibration
6 Conclusion
(1) A ball screw actuator andMR damper are introducedinto vehicle suspension system and a new kind ofhybrid active suspension structure is put forwarde ball screw actuator is prototyped and thefunctional relationship between the back-EMF co-ecient the electromagnetic torque coecient of themotor and the suspension vibration speed is ob-tained by test analyses And the active output me-chanical properties of the ball screw actuator aretested and the results show that the actuator hasgood active output force characteristics
(2) e inshyuences of the variable damping value of thesuspension system on the riding comfort handlingand stability and energy consumption characteristicsof the hybrid suspension in the active control modeare analyzeden the optimal damping values of thehybrid suspension at dierent vehicle speeds aredesigned e eects of electromagnetic dampingforce on the actual semiactive force and the systemenergy of the suspension in the semiactive controlmode are analyzed and then the hybrid suspensionwith semiactive mode which has electromagneticdamping force feedback adjustment is designed Onthis basis a multimode coordinated control strategyfor the hybrid suspension is designed
(3) e damping performance and energy consumptioncharacteristics of the hybrid suspension under cyclicdriving condition are simulated by MATLABSimulink software and the results show that whenin the urban circulation unit compared with activesuspension aw and DTLrms of the hybrid suspensionare reduced by 1322 and 840 respectively Andwhen in the suburban circulation unit comparedwith the active suspension aw and DTLrms of thehybrid suspension are reduced by 1199 and 732respectively e pure energy of the active suspen-sion system during the entire operation cycle is
minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid sus-pension system basically realizes energy self-poweredin theory
(4) e eectiveness verication test of the hybrid sus-pension in active mode and semiactive mode controlis carried out and the results show that when thehybrid suspension is actively controlled comparedwith active suspension aw of the hybrid suspension inactive mode is reduced by 1432 When the hybridsuspension is semiactively controlled compared withsemiactive suspension aw of the hybrid suspension inactive mode is reduced by 307 e test and sim-ulation results are basically consistent and the testveries the correctness of the simulation
Data Availability
e data used to support the ndings of this study are in-cluded within the supplementary information les And thedata used to support the ndings of this study are availablefrom the corresponding author upon request
Conflicts of Interest
e authors declare that there are no conshyicts of interestregarding the publication of this paper
Acknowledgments
is work was supported by the National Natural ScienceFoundation of China (Grant no 51775426) Service LocalSpecial Program Support Project of Shaanxi ProvincialEducation Department (Grant no 17JF017) and XirsquoanScience and Technology Program Funding Project (Grantno 2017079CGRC042-XAKD007)
Supplementary Materials
e supplementary materials are this articlersquos experimentaldata including (1) MR damperrsquos characteristic test data (2)the motor counter electromotive forcersquos peak test data (3)the ball screw actuatorrsquos active output force test data (4) thehybrid suspensionrsquos sprung mass acceleration test data inactive mode (5) the hybrid suspensionrsquos sprung mass ac-celeration test data in semiactive mode and (6) the hybridsuspensionrsquos energy consumption power and energy re-generation power test data (Supplementary Materials)
References
[1] M Montazeri-Gh and O Kavianipour ldquoInvestigation of theactive electromagnetic suspension system considering hybridcontrol strategyrdquo Proceedings of the Institution of MechanicalEngineers Part C Journal of Mechanical Engineering Sciencevol 228 no 10 pp 1658ndash1669 2013
[2] H E Tseng and D Hrovat ldquoState of the art survey active andsemi-active suspension controlrdquo Vehicle System Dynamicsvol 53 no 7 pp 1034ndash1062 2015
0 1 2 3 4 50
20
40
60
80
Time (s)
Pow
er (W
)
Energy consumption power in active modeEnergy regenerative power in semiactive modeEnergy consumption power in semiactive mode
Figure 28 e energy consumption power and energy-re-generative power of the hybrid suspension
Shock and Vibration 15
[3] F Kou J Du Z Wang D Li and J Xu ldquoNonlinear modelingand coordinate optimization of a semi-active energy re-generative suspension with an electro-hydraulic actuatorrdquoChina Mechanical Engineering vol 28 no 14 pp 1701ndash17072017
[4] S H Zareh A Sarrafan A A A Khayyat and A ZabihollahldquoIntelligent semi-active vibration control of eleven degrees offreedom suspension system using magnetorheologicaldampersrdquo Journal of Mechanical Science and Technologyvol 26 no 3 pp 323ndash334 2012
[5] S A Chen X Li L J Zhao Y X Wang and Y B KimldquoDevelopment of a control method for an electromagneticsemi-active suspension reclaiming energy with varying chargevoltage in stepsrdquo International Journal of Automotive Tech-nology vol 16 no 5 pp 765ndash773 2015
[6] K Nakano Y Suda and S Nakadai ldquoSelf-powered activevibration control using a single electric actuatorrdquo Journal ofSound and Vibration vol 260 no 2 pp 213ndash235 2003
[7] K Nakano ldquoCombined type self-powered active vibrationcontrol of truck cabinsrdquo Vehicle System Dynamics vol 41no 6 pp 449ndash473 2004
[8] K Huang Y C Zhang F Yu and Y H Gu ldquoCoordinateoptimization for synthetical performance of electrical energy-regenerative active suspensionrdquo Journal of Shanghai JiaotongUniversity vol 43 no 2 pp 226ndash230 2009
[9] K Huang F Yu and Y C Zhang ldquoActive control of energy-regenerative electromagnetic suspension based on energy flowanalysisrdquo Journal of Shanghai Jiaotong University vol 45no 67 pp 1068ndash1073 2011
[10] D S Huang J Q Zhang Y L Liu L Yi and X Y WangldquoPerformance of a novel energy-regenerative active suspen-sion systemrdquo Journal of Chongqing University (English Edi-tion) vol 14 no 3 pp 109ndash118 2015
[11] B Ebrahimi H Bolandhemmat M B Khamesee andF Golnaraghi ldquoA hybrid electromagnetic shock absorber foractive vehicle suspension systemsrdquo Vehicle System Dynamicsvol 49 no 1-2 pp 311ndash332 2011
[12] S Tang L Chen R Wang X Sun and D Shi ldquoResearch onoptimal control of active suspension based on damping multi-modal switchingrdquo Journal of Guangxi University (NaturalScience) vol 39 no 2 pp 300ndash307 2014
[13] R Wang X Ma R Ding X Meng and L Chen ldquoResearchof multi-mode switching control system for hybrid sus-pension based on model referencerdquo Transactions of theChinese Society for Agricultural Machinery vol 48 no 7pp 353ndash360 2017
[14] R Wang Y Qian R Ding X Meng and J Xie ldquoDesign andtests for damping-stiffness of a hybrid electromagnetic sus-pension based on LQGrdquo Journal of Vibration and Shockvol 37 no 3 pp 61ndash65 2017
[15] B Vanavil K K Chaitanya and A S Rao ldquoImproved PIDcontroller design for unstable time delay processes based ondirect synthesis method and maximum sensitivityrdquo Taylorand Francis vol 46 no 8 pp 1349ndash1366 2015
[16] I Mihai and F Andronic ldquoBehavior of a semi-active sus-pension system versus a passive suspension system on anuneven road surfacerdquo Mechanics vol 20 no 1 pp 64ndash692014
[17] D Ngoduy ldquoLinear stability of a generalized multi-anticipative car following model with time delaysrdquo Com-munications in Nonlinear Science and Numerical Simulationvol 22 no 1 pp 420ndash426 2015
[18] H Zhang E Wang F Min R Subash and C Su ldquoSkyhook-based semi-active control of full-vehicle suspension with
magneto-rheological dampersrdquo Chinese Journal of Mechan-ical Engineering vol 26 no 3 pp 498ndash505 2013
[19] F Kou ldquoAn experimental study on the dynamic character-istics of vehicle semi-active seat suspension with magneto-rheological damperrdquo Automotive Engineering vol 37 no 11pp 1346ndash1352 2015
[20] Z Feng S Chen and Y Liang ldquoAn experimental study on thedynamic characteristics of a megneto-rheological semi-activesuspensionrdquoAutomotive Engineering vol 35 no 1 pp 72ndash772013
[21] Z Li and L Xu A New Type of Magnetorheological Damperand Semi-Active coNtrol Design-eory Science Press BeijingChina 2012
[22] B L J Gysen J J H Paulides J L G Janssen andE A Lomonova ldquoActive electromagnetic suspension systemfor improved vehicle dynamicsrdquo IEEE Transactions on Ve-hicular Technology vol 59 no 3 pp 1156ndash1163 2010
[23] Q N Wang S S Liu W H Wang and H Wei ldquoStructuredesign and parameter matching of ball-screw regenerativedamperrdquo Journal of Jilin University (Engineering and Tech-nology Edition) vol 42 no 5 pp 1100ndash1106 2012
[24] S Chen R He and S Lu ldquoEvaluating system of reclaimingenergy suspension comprehensive performancerdquo Trans-actions of the Chinese Society for Agricultural Machineryvol 37 no 7 pp 14ndash18 2006
[25] Y Fan Control Research on Vehicle Suspension with Electro-Hydrostatic Actuator Xirsquoan University of Science and Tech-nology Xirsquoan China 2017
[26] Z Yu Automobile -eory China Machine Press BeijingChina 2009
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Navigation and Observation
International Journal of
Hindawi
wwwhindawicom Volume 2018
Advances in
Multimedia
Submit your manuscripts atwwwhindawicom
From equation (33) when the hybrid suspension sem-iactive controlled the curves of the total system energychange over time are shown in Figure 17
From Figure 17 when there is the nonfeedback semiactivecontrol the total system energy of the hybrid suspension is60 J And when there is the feedback semiactive control thetotal system energy of the hybrid suspension is 307 J
43 -e Design of Multimode Coordination ControllerWhen the hybrid suspension is actively controlled thesuspension has good vibration isolation performance buthigh energy consumption And when the hybrid suspensionis semiactively controlled the suspension has good eco-nomic performance but the control has limitations Con-sidering that the ball screw actuator can realize active controlof the hybrid suspension in any suspension state a multi-mode coordinated control strategy of the hybrid suspensionis designed
minus csky middot _x2 Fs1113868111386811138681113868
1113868111386811138681113868gt csky middot _x2
11138681113868111386811138681113868
11138681113868111386811138681113868
⎧⎪⎨
⎪⎩
cs c0
⎧⎪⎪⎪⎨
⎪⎪⎪⎩
_x2 minus _x1 1113857 _x2 gt 0
F Fz minus csky middot _x2
cs c0 + ck1113896 _x2 minus _x1 1113857 _x2 le 0
⎧⎪⎪⎪⎪⎪⎪⎪⎪⎪⎨
⎪⎪⎪⎪⎪⎪⎪⎪⎪⎩
(35)
0e frame diagram of the multimode coordinated controlstrategy of the hybrid suspension is shown in Figure 18
From equations (9) (12) (26) and (29) the total systemenergy of the hybrid suspension in multimode coordinatedcontrol is
W2 Wk minusWb minusWZ minusWc (36)
whereW2 is the total system energy of the hybrid suspensionin multimode coordinated control
A hybrid active suspension simulation model is estab-lished by using MATLABSimulink software From Fig-ure 11 when the hybrid suspension is actively controlled theenergy consumption of the MR damper is mainly affected bythe vehicle speed In order to verify the vibration isolationperformance and energy consumption performance of thehybrid active suspension in the cyclic driving conditionsthis article simulates vehicle urban and suburban conditionswhich is based on GBT 19233-2003 ldquoLight Vehicle FuelConsumption Test Methodrdquo among them the urban con-ditions include four cycle units each cycle time is 195 s thesuburban conditions include one cycle unit and the cycletime is 400 s [26] And the schematic diagram of the urbanconditions unit is shown in Figure 19
In order to simulate the vehicle acceleration signal avehicle speed variation model in different cycle units isestablished by using the signal builder function module inSimulink software On this basis the vehicle accelerationmodel in different cycle units is obtained and the vehicleacceleration model can be used as the switching controlmodel for the variable damping when the hybrid suspensionis actively controlled 0e shock absorbers of active sus-pension and passive suspension adopt the original damper ofvehicle and the damping value c1 is 1600 Nmiddotsm Howeverbecause of the long simulation time the dynamic responsecurves of the hybrid active suspension in urban and sub-urban conditions cannot be displayed Figure 20 shows thedynamic response curve of the hybrid active suspension inthe 145ndash175 s in the urban circulation unit among themvehicle in the 145ndash155 s is in a constant speed vehicle in the155ndash163 s is in the deceleration state and vehicle in the163ndash175 s is in other constant speed
0e dynamic response RMS values of the hybrid sus-pension in urban and suburban circulation units are shownin Tables 3 and 4 respectively
From Figure 20 and Tables 3 and 4 the dynamic re-sponses are good when the hybrid suspension is multimodecoordinated control When in the urban circulation unitcompared with the passive suspension aw SWSrms andDTLrms of the hybrid suspension are reduced by 3943225 and 2081 respectively Compared with the activesuspension aw and DTLrms of the hybrid suspension arereduced by 1322 and 840 respectively but SWSrms isincreased by 805 When in the suburban circulation unitcompared with the passive suspension aw SWSrms andDTLrms of the hybrid suspension are reduced by 3916314 and 1955 respectively Compared with the activesuspension aw and DTLrms of the hybrid suspension arereduced by 1199 and 732 respectively but SWSrms isincreased by 845
0e system energy of the hybrid suspension in urban andsuburban circulation units are shown in Figures 21 and 22respectively
Table 2 0e response root mean square values of the hybridsuspension in semiactive control
Without feedback semiactive controlWith feedback semiactive control
Figure 17 0e system energy of the hybrid suspension in semi-active control
10 Shock and Vibration
en the system energy values of the hybrid suspensionin urban and suburban circulation units are shown inTable 5
From Table 5 when in the urban circulation unit thesystem energy of the active suspension and hybrid sus-pension is minus2165 J and 38 J respectively And when in thesuburban circulation unit the system energy of the activesuspension and hybrid suspension is minus15071 J and minus122 Jrespectively e entire operation cycle includes 4 urbancycle units and 1 suburban cycle unit so the pure energy ofthe active suspension system during the entire operationcycle is minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid suspensionsystem basically realizes energy self-powered in theory
5 Test and Analysis
In order to verify the damping eect of the hybrid sus-pension system in active mode and semiactive mode ahybrid suspension vibration test system is designed as
Suspension controllerSpeed of sprung mass x2
Speed of unsprung mass x1Vehicle acceleration v
Vehicle suspension
Ball screw actuator
Output Fk
Active control force Fz
Ball screw actuator
Output Iz
Output Ik
Electromagnetic damping force Fs
Ideal semiactive control force FL
Output semiactive controlforce Fb = Fs
Output semiactive controlforce Fb = FL = Fk + Fs
x2 ndash x1
(x2 ndash x1) x2 gt 0
(x2 ndash x1) x2 le 0
Output Ik
MR damper
Adjust the duty cycle in real timeaccording to the relationshipbetween active output force
and duty cycle
No controllablecurrent output
MR damper
v ne 0
According to vand ∆v
v = 0
MR damper
Adjustable damping cs = c0 + ck
ndashcsky middot x2
ndashcsky middot x2
|Fs| ge |FL|
|Fs| lt |FL|
Figure 18 e frame diagram of the multimode coordinated control strategy of the hybrid suspension
Vehicle shifting
00
15
10
30
30
20
45
40
50
60
60
75 90 105 120Time (s)
Veh
icle
spee
d (k
mh
)
135 150 165 180 195
Figure 19 e schematic diagram of the urban conditions unit
Shock and Vibration 11
shown in Figure 23 During the test the MR damper has nocontrollable current input in the active control mode of thehybrid suspension and its damping value is always 800Nmiddotsm
e passive suspension semiactive suspension and activesuspension all adopt the original damper of the vehicle andthe damping value c1 is 1600Nmiddotsm
Figure 20e dynamic responses of the hybrid suspension (a)e response curves of sprungmass acceleration (b)e response curves ofsuspension working space (c) e response curves of dynamic tire load
12 Shock and Vibration
Because of the limitation of test conditions only thesprung mass acceleration dynamic response of the hybridsuspension is measured in this test e dynamic response ofthe sprung mass acceleration of the hybrid suspension inactive mode under random road is shown in Figure 24 epower spectrum of the sprung mass acceleration of thehybrid suspension in active mode is shown in Figure 25
Table 6 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in active mode
From Table 6 compared with passive suspension aw ofthe hybrid suspension in active mode is reduced by 3945and compared with active suspension aw of the hybridsuspension in active mode is reduced by 1432 And fromFigure 25 when the hybrid suspension is actively controlledcompared with passive suspension the vibration isolationperformance of the suspension in the low frequency andlow-frequency resonance areas is similar to that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withactive suspension the vibration isolation performance of thesuspension in the low frequency high frequency and res-onance regions is slightly better
e dynamic response of the sprung mass acceleration ofthe hybrid suspension in semiactive mode under randomroad is shown in Figure 26 e power spectrum of thesprung mass acceleration of the hybrid suspension insemiactive mode is shown in Figure 27
Table 7 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in semiactive mode
From Table 7 compared with passive suspension aw ofthe hybrid suspension in semiactive mode is reduced by1642 Compared with semiactive suspension aw of thehybrid suspension in active mode is reduced by 307 Andfrom Figure 27 when the hybrid suspension is semiactivelycontrolled compared with passive suspension the vibration
Table 3 e dynamic response root mean square values of thehybrid suspension in urban circulation unit
Figure 21 e system energy of the hybrid suspension in urbancirculation units
0 100 200 300 400ndash20000
ndash15000
ndash10000
ndash5000
0
5000
Time (s)
Susp
ensio
n sy
stem
ener
gy (J
)
Active suspensionHybrid suspension
Figure 22e system energy of the hybrid suspension in suburbancirculation units
Table 5 e system energy values of the hybrid suspension (J)
Cycle unit Active suspension Hybrid suspensionUrban minus2165 38Suburban minus15071 minus122
Figure 23 e test system of the hybrid active suspension
Shock and Vibration 13
isolation performance of the suspension in the low frequencyand low-frequency resonance areas is worse than that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withsemiactive suspension the vibration isolation performance ofthe suspension in the low frequency high frequency andresonance regions is slightly better
e energy consumption power and energy-regenerativepower of the hybrid suspension are shown in Figure 28
From Figure 28 in the active mode the average powerconsumption of the ball screw actuator is 2561W In thesemiactive mode the average regenerative power of the ballscrew actuator is 2696W and the average of theMR damperenergy consumption power is 178W erefore the pureaverage power consumption of the hybrid suspension isminus043W and the test results are basically consistent with thesimulation
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spr
ung
mas
s acc
eler
atio
n (m
s2 )
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 24 e dynamic response of the sprung mass accelerationof the hybrid suspension in active mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 25 e power spectrum of the sprung mass acceleration ofthe hybrid suspension in active mode
Table 6 e sprung mass acceleration RMS of the hybrid sus-pension in active mode
Indicators Passivesuspension
Activesuspension
Hybridsuspension
aw (ms2) 13378 09423 08074
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spru
ng m
ass a
ccel
erat
ion
(ms
2 )
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 26 e dynamic responses of the sprung mass accelerationof the hybrid suspension in semiactive mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 27e power spectrums of the sprungmass acceleration ofthe hybrid suspension in semiactive mode
Table 7 e sprung mass acceleration RMS of the hybrid sus-pension in semiactive mode
Indicators Passivesuspension
Semiactivesuspension
Hybridsuspension
aw (ms2) 13378 11536 11182
14 Shock and Vibration
6 Conclusion
(1) A ball screw actuator andMR damper are introducedinto vehicle suspension system and a new kind ofhybrid active suspension structure is put forwarde ball screw actuator is prototyped and thefunctional relationship between the back-EMF co-ecient the electromagnetic torque coecient of themotor and the suspension vibration speed is ob-tained by test analyses And the active output me-chanical properties of the ball screw actuator aretested and the results show that the actuator hasgood active output force characteristics
(2) e inshyuences of the variable damping value of thesuspension system on the riding comfort handlingand stability and energy consumption characteristicsof the hybrid suspension in the active control modeare analyzeden the optimal damping values of thehybrid suspension at dierent vehicle speeds aredesigned e eects of electromagnetic dampingforce on the actual semiactive force and the systemenergy of the suspension in the semiactive controlmode are analyzed and then the hybrid suspensionwith semiactive mode which has electromagneticdamping force feedback adjustment is designed Onthis basis a multimode coordinated control strategyfor the hybrid suspension is designed
(3) e damping performance and energy consumptioncharacteristics of the hybrid suspension under cyclicdriving condition are simulated by MATLABSimulink software and the results show that whenin the urban circulation unit compared with activesuspension aw and DTLrms of the hybrid suspensionare reduced by 1322 and 840 respectively Andwhen in the suburban circulation unit comparedwith the active suspension aw and DTLrms of thehybrid suspension are reduced by 1199 and 732respectively e pure energy of the active suspen-sion system during the entire operation cycle is
minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid sus-pension system basically realizes energy self-poweredin theory
(4) e eectiveness verication test of the hybrid sus-pension in active mode and semiactive mode controlis carried out and the results show that when thehybrid suspension is actively controlled comparedwith active suspension aw of the hybrid suspension inactive mode is reduced by 1432 When the hybridsuspension is semiactively controlled compared withsemiactive suspension aw of the hybrid suspension inactive mode is reduced by 307 e test and sim-ulation results are basically consistent and the testveries the correctness of the simulation
Data Availability
e data used to support the ndings of this study are in-cluded within the supplementary information les And thedata used to support the ndings of this study are availablefrom the corresponding author upon request
Conflicts of Interest
e authors declare that there are no conshyicts of interestregarding the publication of this paper
Acknowledgments
is work was supported by the National Natural ScienceFoundation of China (Grant no 51775426) Service LocalSpecial Program Support Project of Shaanxi ProvincialEducation Department (Grant no 17JF017) and XirsquoanScience and Technology Program Funding Project (Grantno 2017079CGRC042-XAKD007)
Supplementary Materials
e supplementary materials are this articlersquos experimentaldata including (1) MR damperrsquos characteristic test data (2)the motor counter electromotive forcersquos peak test data (3)the ball screw actuatorrsquos active output force test data (4) thehybrid suspensionrsquos sprung mass acceleration test data inactive mode (5) the hybrid suspensionrsquos sprung mass ac-celeration test data in semiactive mode and (6) the hybridsuspensionrsquos energy consumption power and energy re-generation power test data (Supplementary Materials)
References
[1] M Montazeri-Gh and O Kavianipour ldquoInvestigation of theactive electromagnetic suspension system considering hybridcontrol strategyrdquo Proceedings of the Institution of MechanicalEngineers Part C Journal of Mechanical Engineering Sciencevol 228 no 10 pp 1658ndash1669 2013
[2] H E Tseng and D Hrovat ldquoState of the art survey active andsemi-active suspension controlrdquo Vehicle System Dynamicsvol 53 no 7 pp 1034ndash1062 2015
0 1 2 3 4 50
20
40
60
80
Time (s)
Pow
er (W
)
Energy consumption power in active modeEnergy regenerative power in semiactive modeEnergy consumption power in semiactive mode
Figure 28 e energy consumption power and energy-re-generative power of the hybrid suspension
Shock and Vibration 15
[3] F Kou J Du Z Wang D Li and J Xu ldquoNonlinear modelingand coordinate optimization of a semi-active energy re-generative suspension with an electro-hydraulic actuatorrdquoChina Mechanical Engineering vol 28 no 14 pp 1701ndash17072017
[4] S H Zareh A Sarrafan A A A Khayyat and A ZabihollahldquoIntelligent semi-active vibration control of eleven degrees offreedom suspension system using magnetorheologicaldampersrdquo Journal of Mechanical Science and Technologyvol 26 no 3 pp 323ndash334 2012
[5] S A Chen X Li L J Zhao Y X Wang and Y B KimldquoDevelopment of a control method for an electromagneticsemi-active suspension reclaiming energy with varying chargevoltage in stepsrdquo International Journal of Automotive Tech-nology vol 16 no 5 pp 765ndash773 2015
[6] K Nakano Y Suda and S Nakadai ldquoSelf-powered activevibration control using a single electric actuatorrdquo Journal ofSound and Vibration vol 260 no 2 pp 213ndash235 2003
[7] K Nakano ldquoCombined type self-powered active vibrationcontrol of truck cabinsrdquo Vehicle System Dynamics vol 41no 6 pp 449ndash473 2004
[8] K Huang Y C Zhang F Yu and Y H Gu ldquoCoordinateoptimization for synthetical performance of electrical energy-regenerative active suspensionrdquo Journal of Shanghai JiaotongUniversity vol 43 no 2 pp 226ndash230 2009
[9] K Huang F Yu and Y C Zhang ldquoActive control of energy-regenerative electromagnetic suspension based on energy flowanalysisrdquo Journal of Shanghai Jiaotong University vol 45no 67 pp 1068ndash1073 2011
[10] D S Huang J Q Zhang Y L Liu L Yi and X Y WangldquoPerformance of a novel energy-regenerative active suspen-sion systemrdquo Journal of Chongqing University (English Edi-tion) vol 14 no 3 pp 109ndash118 2015
[11] B Ebrahimi H Bolandhemmat M B Khamesee andF Golnaraghi ldquoA hybrid electromagnetic shock absorber foractive vehicle suspension systemsrdquo Vehicle System Dynamicsvol 49 no 1-2 pp 311ndash332 2011
[12] S Tang L Chen R Wang X Sun and D Shi ldquoResearch onoptimal control of active suspension based on damping multi-modal switchingrdquo Journal of Guangxi University (NaturalScience) vol 39 no 2 pp 300ndash307 2014
[13] R Wang X Ma R Ding X Meng and L Chen ldquoResearchof multi-mode switching control system for hybrid sus-pension based on model referencerdquo Transactions of theChinese Society for Agricultural Machinery vol 48 no 7pp 353ndash360 2017
[14] R Wang Y Qian R Ding X Meng and J Xie ldquoDesign andtests for damping-stiffness of a hybrid electromagnetic sus-pension based on LQGrdquo Journal of Vibration and Shockvol 37 no 3 pp 61ndash65 2017
[15] B Vanavil K K Chaitanya and A S Rao ldquoImproved PIDcontroller design for unstable time delay processes based ondirect synthesis method and maximum sensitivityrdquo Taylorand Francis vol 46 no 8 pp 1349ndash1366 2015
[16] I Mihai and F Andronic ldquoBehavior of a semi-active sus-pension system versus a passive suspension system on anuneven road surfacerdquo Mechanics vol 20 no 1 pp 64ndash692014
[17] D Ngoduy ldquoLinear stability of a generalized multi-anticipative car following model with time delaysrdquo Com-munications in Nonlinear Science and Numerical Simulationvol 22 no 1 pp 420ndash426 2015
[18] H Zhang E Wang F Min R Subash and C Su ldquoSkyhook-based semi-active control of full-vehicle suspension with
magneto-rheological dampersrdquo Chinese Journal of Mechan-ical Engineering vol 26 no 3 pp 498ndash505 2013
[19] F Kou ldquoAn experimental study on the dynamic character-istics of vehicle semi-active seat suspension with magneto-rheological damperrdquo Automotive Engineering vol 37 no 11pp 1346ndash1352 2015
[20] Z Feng S Chen and Y Liang ldquoAn experimental study on thedynamic characteristics of a megneto-rheological semi-activesuspensionrdquoAutomotive Engineering vol 35 no 1 pp 72ndash772013
[21] Z Li and L Xu A New Type of Magnetorheological Damperand Semi-Active coNtrol Design-eory Science Press BeijingChina 2012
[22] B L J Gysen J J H Paulides J L G Janssen andE A Lomonova ldquoActive electromagnetic suspension systemfor improved vehicle dynamicsrdquo IEEE Transactions on Ve-hicular Technology vol 59 no 3 pp 1156ndash1163 2010
[23] Q N Wang S S Liu W H Wang and H Wei ldquoStructuredesign and parameter matching of ball-screw regenerativedamperrdquo Journal of Jilin University (Engineering and Tech-nology Edition) vol 42 no 5 pp 1100ndash1106 2012
[24] S Chen R He and S Lu ldquoEvaluating system of reclaimingenergy suspension comprehensive performancerdquo Trans-actions of the Chinese Society for Agricultural Machineryvol 37 no 7 pp 14ndash18 2006
[25] Y Fan Control Research on Vehicle Suspension with Electro-Hydrostatic Actuator Xirsquoan University of Science and Tech-nology Xirsquoan China 2017
[26] Z Yu Automobile -eory China Machine Press BeijingChina 2009
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Navigation and Observation
International Journal of
Hindawi
wwwhindawicom Volume 2018
Advances in
Multimedia
Submit your manuscripts atwwwhindawicom
en the system energy values of the hybrid suspensionin urban and suburban circulation units are shown inTable 5
From Table 5 when in the urban circulation unit thesystem energy of the active suspension and hybrid sus-pension is minus2165 J and 38 J respectively And when in thesuburban circulation unit the system energy of the activesuspension and hybrid suspension is minus15071 J and minus122 Jrespectively e entire operation cycle includes 4 urbancycle units and 1 suburban cycle unit so the pure energy ofthe active suspension system during the entire operationcycle is minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid suspensionsystem basically realizes energy self-powered in theory
5 Test and Analysis
In order to verify the damping eect of the hybrid sus-pension system in active mode and semiactive mode ahybrid suspension vibration test system is designed as
Suspension controllerSpeed of sprung mass x2
Speed of unsprung mass x1Vehicle acceleration v
Vehicle suspension
Ball screw actuator
Output Fk
Active control force Fz
Ball screw actuator
Output Iz
Output Ik
Electromagnetic damping force Fs
Ideal semiactive control force FL
Output semiactive controlforce Fb = Fs
Output semiactive controlforce Fb = FL = Fk + Fs
x2 ndash x1
(x2 ndash x1) x2 gt 0
(x2 ndash x1) x2 le 0
Output Ik
MR damper
Adjust the duty cycle in real timeaccording to the relationshipbetween active output force
and duty cycle
No controllablecurrent output
MR damper
v ne 0
According to vand ∆v
v = 0
MR damper
Adjustable damping cs = c0 + ck
ndashcsky middot x2
ndashcsky middot x2
|Fs| ge |FL|
|Fs| lt |FL|
Figure 18 e frame diagram of the multimode coordinated control strategy of the hybrid suspension
Vehicle shifting
00
15
10
30
30
20
45
40
50
60
60
75 90 105 120Time (s)
Veh
icle
spee
d (k
mh
)
135 150 165 180 195
Figure 19 e schematic diagram of the urban conditions unit
Shock and Vibration 11
shown in Figure 23 During the test the MR damper has nocontrollable current input in the active control mode of thehybrid suspension and its damping value is always 800Nmiddotsm
e passive suspension semiactive suspension and activesuspension all adopt the original damper of the vehicle andthe damping value c1 is 1600Nmiddotsm
Figure 20e dynamic responses of the hybrid suspension (a)e response curves of sprungmass acceleration (b)e response curves ofsuspension working space (c) e response curves of dynamic tire load
12 Shock and Vibration
Because of the limitation of test conditions only thesprung mass acceleration dynamic response of the hybridsuspension is measured in this test e dynamic response ofthe sprung mass acceleration of the hybrid suspension inactive mode under random road is shown in Figure 24 epower spectrum of the sprung mass acceleration of thehybrid suspension in active mode is shown in Figure 25
Table 6 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in active mode
From Table 6 compared with passive suspension aw ofthe hybrid suspension in active mode is reduced by 3945and compared with active suspension aw of the hybridsuspension in active mode is reduced by 1432 And fromFigure 25 when the hybrid suspension is actively controlledcompared with passive suspension the vibration isolationperformance of the suspension in the low frequency andlow-frequency resonance areas is similar to that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withactive suspension the vibration isolation performance of thesuspension in the low frequency high frequency and res-onance regions is slightly better
e dynamic response of the sprung mass acceleration ofthe hybrid suspension in semiactive mode under randomroad is shown in Figure 26 e power spectrum of thesprung mass acceleration of the hybrid suspension insemiactive mode is shown in Figure 27
Table 7 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in semiactive mode
From Table 7 compared with passive suspension aw ofthe hybrid suspension in semiactive mode is reduced by1642 Compared with semiactive suspension aw of thehybrid suspension in active mode is reduced by 307 Andfrom Figure 27 when the hybrid suspension is semiactivelycontrolled compared with passive suspension the vibration
Table 3 e dynamic response root mean square values of thehybrid suspension in urban circulation unit
Figure 21 e system energy of the hybrid suspension in urbancirculation units
0 100 200 300 400ndash20000
ndash15000
ndash10000
ndash5000
0
5000
Time (s)
Susp
ensio
n sy
stem
ener
gy (J
)
Active suspensionHybrid suspension
Figure 22e system energy of the hybrid suspension in suburbancirculation units
Table 5 e system energy values of the hybrid suspension (J)
Cycle unit Active suspension Hybrid suspensionUrban minus2165 38Suburban minus15071 minus122
Figure 23 e test system of the hybrid active suspension
Shock and Vibration 13
isolation performance of the suspension in the low frequencyand low-frequency resonance areas is worse than that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withsemiactive suspension the vibration isolation performance ofthe suspension in the low frequency high frequency andresonance regions is slightly better
e energy consumption power and energy-regenerativepower of the hybrid suspension are shown in Figure 28
From Figure 28 in the active mode the average powerconsumption of the ball screw actuator is 2561W In thesemiactive mode the average regenerative power of the ballscrew actuator is 2696W and the average of theMR damperenergy consumption power is 178W erefore the pureaverage power consumption of the hybrid suspension isminus043W and the test results are basically consistent with thesimulation
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spr
ung
mas
s acc
eler
atio
n (m
s2 )
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 24 e dynamic response of the sprung mass accelerationof the hybrid suspension in active mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 25 e power spectrum of the sprung mass acceleration ofthe hybrid suspension in active mode
Table 6 e sprung mass acceleration RMS of the hybrid sus-pension in active mode
Indicators Passivesuspension
Activesuspension
Hybridsuspension
aw (ms2) 13378 09423 08074
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spru
ng m
ass a
ccel
erat
ion
(ms
2 )
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 26 e dynamic responses of the sprung mass accelerationof the hybrid suspension in semiactive mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 27e power spectrums of the sprungmass acceleration ofthe hybrid suspension in semiactive mode
Table 7 e sprung mass acceleration RMS of the hybrid sus-pension in semiactive mode
Indicators Passivesuspension
Semiactivesuspension
Hybridsuspension
aw (ms2) 13378 11536 11182
14 Shock and Vibration
6 Conclusion
(1) A ball screw actuator andMR damper are introducedinto vehicle suspension system and a new kind ofhybrid active suspension structure is put forwarde ball screw actuator is prototyped and thefunctional relationship between the back-EMF co-ecient the electromagnetic torque coecient of themotor and the suspension vibration speed is ob-tained by test analyses And the active output me-chanical properties of the ball screw actuator aretested and the results show that the actuator hasgood active output force characteristics
(2) e inshyuences of the variable damping value of thesuspension system on the riding comfort handlingand stability and energy consumption characteristicsof the hybrid suspension in the active control modeare analyzeden the optimal damping values of thehybrid suspension at dierent vehicle speeds aredesigned e eects of electromagnetic dampingforce on the actual semiactive force and the systemenergy of the suspension in the semiactive controlmode are analyzed and then the hybrid suspensionwith semiactive mode which has electromagneticdamping force feedback adjustment is designed Onthis basis a multimode coordinated control strategyfor the hybrid suspension is designed
(3) e damping performance and energy consumptioncharacteristics of the hybrid suspension under cyclicdriving condition are simulated by MATLABSimulink software and the results show that whenin the urban circulation unit compared with activesuspension aw and DTLrms of the hybrid suspensionare reduced by 1322 and 840 respectively Andwhen in the suburban circulation unit comparedwith the active suspension aw and DTLrms of thehybrid suspension are reduced by 1199 and 732respectively e pure energy of the active suspen-sion system during the entire operation cycle is
minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid sus-pension system basically realizes energy self-poweredin theory
(4) e eectiveness verication test of the hybrid sus-pension in active mode and semiactive mode controlis carried out and the results show that when thehybrid suspension is actively controlled comparedwith active suspension aw of the hybrid suspension inactive mode is reduced by 1432 When the hybridsuspension is semiactively controlled compared withsemiactive suspension aw of the hybrid suspension inactive mode is reduced by 307 e test and sim-ulation results are basically consistent and the testveries the correctness of the simulation
Data Availability
e data used to support the ndings of this study are in-cluded within the supplementary information les And thedata used to support the ndings of this study are availablefrom the corresponding author upon request
Conflicts of Interest
e authors declare that there are no conshyicts of interestregarding the publication of this paper
Acknowledgments
is work was supported by the National Natural ScienceFoundation of China (Grant no 51775426) Service LocalSpecial Program Support Project of Shaanxi ProvincialEducation Department (Grant no 17JF017) and XirsquoanScience and Technology Program Funding Project (Grantno 2017079CGRC042-XAKD007)
Supplementary Materials
e supplementary materials are this articlersquos experimentaldata including (1) MR damperrsquos characteristic test data (2)the motor counter electromotive forcersquos peak test data (3)the ball screw actuatorrsquos active output force test data (4) thehybrid suspensionrsquos sprung mass acceleration test data inactive mode (5) the hybrid suspensionrsquos sprung mass ac-celeration test data in semiactive mode and (6) the hybridsuspensionrsquos energy consumption power and energy re-generation power test data (Supplementary Materials)
References
[1] M Montazeri-Gh and O Kavianipour ldquoInvestigation of theactive electromagnetic suspension system considering hybridcontrol strategyrdquo Proceedings of the Institution of MechanicalEngineers Part C Journal of Mechanical Engineering Sciencevol 228 no 10 pp 1658ndash1669 2013
[2] H E Tseng and D Hrovat ldquoState of the art survey active andsemi-active suspension controlrdquo Vehicle System Dynamicsvol 53 no 7 pp 1034ndash1062 2015
0 1 2 3 4 50
20
40
60
80
Time (s)
Pow
er (W
)
Energy consumption power in active modeEnergy regenerative power in semiactive modeEnergy consumption power in semiactive mode
Figure 28 e energy consumption power and energy-re-generative power of the hybrid suspension
Shock and Vibration 15
[3] F Kou J Du Z Wang D Li and J Xu ldquoNonlinear modelingand coordinate optimization of a semi-active energy re-generative suspension with an electro-hydraulic actuatorrdquoChina Mechanical Engineering vol 28 no 14 pp 1701ndash17072017
[4] S H Zareh A Sarrafan A A A Khayyat and A ZabihollahldquoIntelligent semi-active vibration control of eleven degrees offreedom suspension system using magnetorheologicaldampersrdquo Journal of Mechanical Science and Technologyvol 26 no 3 pp 323ndash334 2012
[5] S A Chen X Li L J Zhao Y X Wang and Y B KimldquoDevelopment of a control method for an electromagneticsemi-active suspension reclaiming energy with varying chargevoltage in stepsrdquo International Journal of Automotive Tech-nology vol 16 no 5 pp 765ndash773 2015
[6] K Nakano Y Suda and S Nakadai ldquoSelf-powered activevibration control using a single electric actuatorrdquo Journal ofSound and Vibration vol 260 no 2 pp 213ndash235 2003
[7] K Nakano ldquoCombined type self-powered active vibrationcontrol of truck cabinsrdquo Vehicle System Dynamics vol 41no 6 pp 449ndash473 2004
[8] K Huang Y C Zhang F Yu and Y H Gu ldquoCoordinateoptimization for synthetical performance of electrical energy-regenerative active suspensionrdquo Journal of Shanghai JiaotongUniversity vol 43 no 2 pp 226ndash230 2009
[9] K Huang F Yu and Y C Zhang ldquoActive control of energy-regenerative electromagnetic suspension based on energy flowanalysisrdquo Journal of Shanghai Jiaotong University vol 45no 67 pp 1068ndash1073 2011
[10] D S Huang J Q Zhang Y L Liu L Yi and X Y WangldquoPerformance of a novel energy-regenerative active suspen-sion systemrdquo Journal of Chongqing University (English Edi-tion) vol 14 no 3 pp 109ndash118 2015
[11] B Ebrahimi H Bolandhemmat M B Khamesee andF Golnaraghi ldquoA hybrid electromagnetic shock absorber foractive vehicle suspension systemsrdquo Vehicle System Dynamicsvol 49 no 1-2 pp 311ndash332 2011
[12] S Tang L Chen R Wang X Sun and D Shi ldquoResearch onoptimal control of active suspension based on damping multi-modal switchingrdquo Journal of Guangxi University (NaturalScience) vol 39 no 2 pp 300ndash307 2014
[13] R Wang X Ma R Ding X Meng and L Chen ldquoResearchof multi-mode switching control system for hybrid sus-pension based on model referencerdquo Transactions of theChinese Society for Agricultural Machinery vol 48 no 7pp 353ndash360 2017
[14] R Wang Y Qian R Ding X Meng and J Xie ldquoDesign andtests for damping-stiffness of a hybrid electromagnetic sus-pension based on LQGrdquo Journal of Vibration and Shockvol 37 no 3 pp 61ndash65 2017
[15] B Vanavil K K Chaitanya and A S Rao ldquoImproved PIDcontroller design for unstable time delay processes based ondirect synthesis method and maximum sensitivityrdquo Taylorand Francis vol 46 no 8 pp 1349ndash1366 2015
[16] I Mihai and F Andronic ldquoBehavior of a semi-active sus-pension system versus a passive suspension system on anuneven road surfacerdquo Mechanics vol 20 no 1 pp 64ndash692014
[17] D Ngoduy ldquoLinear stability of a generalized multi-anticipative car following model with time delaysrdquo Com-munications in Nonlinear Science and Numerical Simulationvol 22 no 1 pp 420ndash426 2015
[18] H Zhang E Wang F Min R Subash and C Su ldquoSkyhook-based semi-active control of full-vehicle suspension with
magneto-rheological dampersrdquo Chinese Journal of Mechan-ical Engineering vol 26 no 3 pp 498ndash505 2013
[19] F Kou ldquoAn experimental study on the dynamic character-istics of vehicle semi-active seat suspension with magneto-rheological damperrdquo Automotive Engineering vol 37 no 11pp 1346ndash1352 2015
[20] Z Feng S Chen and Y Liang ldquoAn experimental study on thedynamic characteristics of a megneto-rheological semi-activesuspensionrdquoAutomotive Engineering vol 35 no 1 pp 72ndash772013
[21] Z Li and L Xu A New Type of Magnetorheological Damperand Semi-Active coNtrol Design-eory Science Press BeijingChina 2012
[22] B L J Gysen J J H Paulides J L G Janssen andE A Lomonova ldquoActive electromagnetic suspension systemfor improved vehicle dynamicsrdquo IEEE Transactions on Ve-hicular Technology vol 59 no 3 pp 1156ndash1163 2010
[23] Q N Wang S S Liu W H Wang and H Wei ldquoStructuredesign and parameter matching of ball-screw regenerativedamperrdquo Journal of Jilin University (Engineering and Tech-nology Edition) vol 42 no 5 pp 1100ndash1106 2012
[24] S Chen R He and S Lu ldquoEvaluating system of reclaimingenergy suspension comprehensive performancerdquo Trans-actions of the Chinese Society for Agricultural Machineryvol 37 no 7 pp 14ndash18 2006
[25] Y Fan Control Research on Vehicle Suspension with Electro-Hydrostatic Actuator Xirsquoan University of Science and Tech-nology Xirsquoan China 2017
[26] Z Yu Automobile -eory China Machine Press BeijingChina 2009
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Navigation and Observation
International Journal of
Hindawi
wwwhindawicom Volume 2018
Advances in
Multimedia
Submit your manuscripts atwwwhindawicom
shown in Figure 23 During the test the MR damper has nocontrollable current input in the active control mode of thehybrid suspension and its damping value is always 800Nmiddotsm
e passive suspension semiactive suspension and activesuspension all adopt the original damper of the vehicle andthe damping value c1 is 1600Nmiddotsm
Figure 20e dynamic responses of the hybrid suspension (a)e response curves of sprungmass acceleration (b)e response curves ofsuspension working space (c) e response curves of dynamic tire load
12 Shock and Vibration
Because of the limitation of test conditions only thesprung mass acceleration dynamic response of the hybridsuspension is measured in this test e dynamic response ofthe sprung mass acceleration of the hybrid suspension inactive mode under random road is shown in Figure 24 epower spectrum of the sprung mass acceleration of thehybrid suspension in active mode is shown in Figure 25
Table 6 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in active mode
From Table 6 compared with passive suspension aw ofthe hybrid suspension in active mode is reduced by 3945and compared with active suspension aw of the hybridsuspension in active mode is reduced by 1432 And fromFigure 25 when the hybrid suspension is actively controlledcompared with passive suspension the vibration isolationperformance of the suspension in the low frequency andlow-frequency resonance areas is similar to that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withactive suspension the vibration isolation performance of thesuspension in the low frequency high frequency and res-onance regions is slightly better
e dynamic response of the sprung mass acceleration ofthe hybrid suspension in semiactive mode under randomroad is shown in Figure 26 e power spectrum of thesprung mass acceleration of the hybrid suspension insemiactive mode is shown in Figure 27
Table 7 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in semiactive mode
From Table 7 compared with passive suspension aw ofthe hybrid suspension in semiactive mode is reduced by1642 Compared with semiactive suspension aw of thehybrid suspension in active mode is reduced by 307 Andfrom Figure 27 when the hybrid suspension is semiactivelycontrolled compared with passive suspension the vibration
Table 3 e dynamic response root mean square values of thehybrid suspension in urban circulation unit
Figure 21 e system energy of the hybrid suspension in urbancirculation units
0 100 200 300 400ndash20000
ndash15000
ndash10000
ndash5000
0
5000
Time (s)
Susp
ensio
n sy
stem
ener
gy (J
)
Active suspensionHybrid suspension
Figure 22e system energy of the hybrid suspension in suburbancirculation units
Table 5 e system energy values of the hybrid suspension (J)
Cycle unit Active suspension Hybrid suspensionUrban minus2165 38Suburban minus15071 minus122
Figure 23 e test system of the hybrid active suspension
Shock and Vibration 13
isolation performance of the suspension in the low frequencyand low-frequency resonance areas is worse than that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withsemiactive suspension the vibration isolation performance ofthe suspension in the low frequency high frequency andresonance regions is slightly better
e energy consumption power and energy-regenerativepower of the hybrid suspension are shown in Figure 28
From Figure 28 in the active mode the average powerconsumption of the ball screw actuator is 2561W In thesemiactive mode the average regenerative power of the ballscrew actuator is 2696W and the average of theMR damperenergy consumption power is 178W erefore the pureaverage power consumption of the hybrid suspension isminus043W and the test results are basically consistent with thesimulation
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spr
ung
mas
s acc
eler
atio
n (m
s2 )
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 24 e dynamic response of the sprung mass accelerationof the hybrid suspension in active mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 25 e power spectrum of the sprung mass acceleration ofthe hybrid suspension in active mode
Table 6 e sprung mass acceleration RMS of the hybrid sus-pension in active mode
Indicators Passivesuspension
Activesuspension
Hybridsuspension
aw (ms2) 13378 09423 08074
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spru
ng m
ass a
ccel
erat
ion
(ms
2 )
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 26 e dynamic responses of the sprung mass accelerationof the hybrid suspension in semiactive mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 27e power spectrums of the sprungmass acceleration ofthe hybrid suspension in semiactive mode
Table 7 e sprung mass acceleration RMS of the hybrid sus-pension in semiactive mode
Indicators Passivesuspension
Semiactivesuspension
Hybridsuspension
aw (ms2) 13378 11536 11182
14 Shock and Vibration
6 Conclusion
(1) A ball screw actuator andMR damper are introducedinto vehicle suspension system and a new kind ofhybrid active suspension structure is put forwarde ball screw actuator is prototyped and thefunctional relationship between the back-EMF co-ecient the electromagnetic torque coecient of themotor and the suspension vibration speed is ob-tained by test analyses And the active output me-chanical properties of the ball screw actuator aretested and the results show that the actuator hasgood active output force characteristics
(2) e inshyuences of the variable damping value of thesuspension system on the riding comfort handlingand stability and energy consumption characteristicsof the hybrid suspension in the active control modeare analyzeden the optimal damping values of thehybrid suspension at dierent vehicle speeds aredesigned e eects of electromagnetic dampingforce on the actual semiactive force and the systemenergy of the suspension in the semiactive controlmode are analyzed and then the hybrid suspensionwith semiactive mode which has electromagneticdamping force feedback adjustment is designed Onthis basis a multimode coordinated control strategyfor the hybrid suspension is designed
(3) e damping performance and energy consumptioncharacteristics of the hybrid suspension under cyclicdriving condition are simulated by MATLABSimulink software and the results show that whenin the urban circulation unit compared with activesuspension aw and DTLrms of the hybrid suspensionare reduced by 1322 and 840 respectively Andwhen in the suburban circulation unit comparedwith the active suspension aw and DTLrms of thehybrid suspension are reduced by 1199 and 732respectively e pure energy of the active suspen-sion system during the entire operation cycle is
minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid sus-pension system basically realizes energy self-poweredin theory
(4) e eectiveness verication test of the hybrid sus-pension in active mode and semiactive mode controlis carried out and the results show that when thehybrid suspension is actively controlled comparedwith active suspension aw of the hybrid suspension inactive mode is reduced by 1432 When the hybridsuspension is semiactively controlled compared withsemiactive suspension aw of the hybrid suspension inactive mode is reduced by 307 e test and sim-ulation results are basically consistent and the testveries the correctness of the simulation
Data Availability
e data used to support the ndings of this study are in-cluded within the supplementary information les And thedata used to support the ndings of this study are availablefrom the corresponding author upon request
Conflicts of Interest
e authors declare that there are no conshyicts of interestregarding the publication of this paper
Acknowledgments
is work was supported by the National Natural ScienceFoundation of China (Grant no 51775426) Service LocalSpecial Program Support Project of Shaanxi ProvincialEducation Department (Grant no 17JF017) and XirsquoanScience and Technology Program Funding Project (Grantno 2017079CGRC042-XAKD007)
Supplementary Materials
e supplementary materials are this articlersquos experimentaldata including (1) MR damperrsquos characteristic test data (2)the motor counter electromotive forcersquos peak test data (3)the ball screw actuatorrsquos active output force test data (4) thehybrid suspensionrsquos sprung mass acceleration test data inactive mode (5) the hybrid suspensionrsquos sprung mass ac-celeration test data in semiactive mode and (6) the hybridsuspensionrsquos energy consumption power and energy re-generation power test data (Supplementary Materials)
References
[1] M Montazeri-Gh and O Kavianipour ldquoInvestigation of theactive electromagnetic suspension system considering hybridcontrol strategyrdquo Proceedings of the Institution of MechanicalEngineers Part C Journal of Mechanical Engineering Sciencevol 228 no 10 pp 1658ndash1669 2013
[2] H E Tseng and D Hrovat ldquoState of the art survey active andsemi-active suspension controlrdquo Vehicle System Dynamicsvol 53 no 7 pp 1034ndash1062 2015
0 1 2 3 4 50
20
40
60
80
Time (s)
Pow
er (W
)
Energy consumption power in active modeEnergy regenerative power in semiactive modeEnergy consumption power in semiactive mode
Figure 28 e energy consumption power and energy-re-generative power of the hybrid suspension
Shock and Vibration 15
[3] F Kou J Du Z Wang D Li and J Xu ldquoNonlinear modelingand coordinate optimization of a semi-active energy re-generative suspension with an electro-hydraulic actuatorrdquoChina Mechanical Engineering vol 28 no 14 pp 1701ndash17072017
[4] S H Zareh A Sarrafan A A A Khayyat and A ZabihollahldquoIntelligent semi-active vibration control of eleven degrees offreedom suspension system using magnetorheologicaldampersrdquo Journal of Mechanical Science and Technologyvol 26 no 3 pp 323ndash334 2012
[5] S A Chen X Li L J Zhao Y X Wang and Y B KimldquoDevelopment of a control method for an electromagneticsemi-active suspension reclaiming energy with varying chargevoltage in stepsrdquo International Journal of Automotive Tech-nology vol 16 no 5 pp 765ndash773 2015
[6] K Nakano Y Suda and S Nakadai ldquoSelf-powered activevibration control using a single electric actuatorrdquo Journal ofSound and Vibration vol 260 no 2 pp 213ndash235 2003
[7] K Nakano ldquoCombined type self-powered active vibrationcontrol of truck cabinsrdquo Vehicle System Dynamics vol 41no 6 pp 449ndash473 2004
[8] K Huang Y C Zhang F Yu and Y H Gu ldquoCoordinateoptimization for synthetical performance of electrical energy-regenerative active suspensionrdquo Journal of Shanghai JiaotongUniversity vol 43 no 2 pp 226ndash230 2009
[9] K Huang F Yu and Y C Zhang ldquoActive control of energy-regenerative electromagnetic suspension based on energy flowanalysisrdquo Journal of Shanghai Jiaotong University vol 45no 67 pp 1068ndash1073 2011
[10] D S Huang J Q Zhang Y L Liu L Yi and X Y WangldquoPerformance of a novel energy-regenerative active suspen-sion systemrdquo Journal of Chongqing University (English Edi-tion) vol 14 no 3 pp 109ndash118 2015
[11] B Ebrahimi H Bolandhemmat M B Khamesee andF Golnaraghi ldquoA hybrid electromagnetic shock absorber foractive vehicle suspension systemsrdquo Vehicle System Dynamicsvol 49 no 1-2 pp 311ndash332 2011
[12] S Tang L Chen R Wang X Sun and D Shi ldquoResearch onoptimal control of active suspension based on damping multi-modal switchingrdquo Journal of Guangxi University (NaturalScience) vol 39 no 2 pp 300ndash307 2014
[13] R Wang X Ma R Ding X Meng and L Chen ldquoResearchof multi-mode switching control system for hybrid sus-pension based on model referencerdquo Transactions of theChinese Society for Agricultural Machinery vol 48 no 7pp 353ndash360 2017
[14] R Wang Y Qian R Ding X Meng and J Xie ldquoDesign andtests for damping-stiffness of a hybrid electromagnetic sus-pension based on LQGrdquo Journal of Vibration and Shockvol 37 no 3 pp 61ndash65 2017
[15] B Vanavil K K Chaitanya and A S Rao ldquoImproved PIDcontroller design for unstable time delay processes based ondirect synthesis method and maximum sensitivityrdquo Taylorand Francis vol 46 no 8 pp 1349ndash1366 2015
[16] I Mihai and F Andronic ldquoBehavior of a semi-active sus-pension system versus a passive suspension system on anuneven road surfacerdquo Mechanics vol 20 no 1 pp 64ndash692014
[17] D Ngoduy ldquoLinear stability of a generalized multi-anticipative car following model with time delaysrdquo Com-munications in Nonlinear Science and Numerical Simulationvol 22 no 1 pp 420ndash426 2015
[18] H Zhang E Wang F Min R Subash and C Su ldquoSkyhook-based semi-active control of full-vehicle suspension with
magneto-rheological dampersrdquo Chinese Journal of Mechan-ical Engineering vol 26 no 3 pp 498ndash505 2013
[19] F Kou ldquoAn experimental study on the dynamic character-istics of vehicle semi-active seat suspension with magneto-rheological damperrdquo Automotive Engineering vol 37 no 11pp 1346ndash1352 2015
[20] Z Feng S Chen and Y Liang ldquoAn experimental study on thedynamic characteristics of a megneto-rheological semi-activesuspensionrdquoAutomotive Engineering vol 35 no 1 pp 72ndash772013
[21] Z Li and L Xu A New Type of Magnetorheological Damperand Semi-Active coNtrol Design-eory Science Press BeijingChina 2012
[22] B L J Gysen J J H Paulides J L G Janssen andE A Lomonova ldquoActive electromagnetic suspension systemfor improved vehicle dynamicsrdquo IEEE Transactions on Ve-hicular Technology vol 59 no 3 pp 1156ndash1163 2010
[23] Q N Wang S S Liu W H Wang and H Wei ldquoStructuredesign and parameter matching of ball-screw regenerativedamperrdquo Journal of Jilin University (Engineering and Tech-nology Edition) vol 42 no 5 pp 1100ndash1106 2012
[24] S Chen R He and S Lu ldquoEvaluating system of reclaimingenergy suspension comprehensive performancerdquo Trans-actions of the Chinese Society for Agricultural Machineryvol 37 no 7 pp 14ndash18 2006
[25] Y Fan Control Research on Vehicle Suspension with Electro-Hydrostatic Actuator Xirsquoan University of Science and Tech-nology Xirsquoan China 2017
[26] Z Yu Automobile -eory China Machine Press BeijingChina 2009
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Navigation and Observation
International Journal of
Hindawi
wwwhindawicom Volume 2018
Advances in
Multimedia
Submit your manuscripts atwwwhindawicom
Because of the limitation of test conditions only thesprung mass acceleration dynamic response of the hybridsuspension is measured in this test e dynamic response ofthe sprung mass acceleration of the hybrid suspension inactive mode under random road is shown in Figure 24 epower spectrum of the sprung mass acceleration of thehybrid suspension in active mode is shown in Figure 25
Table 6 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in active mode
From Table 6 compared with passive suspension aw ofthe hybrid suspension in active mode is reduced by 3945and compared with active suspension aw of the hybridsuspension in active mode is reduced by 1432 And fromFigure 25 when the hybrid suspension is actively controlledcompared with passive suspension the vibration isolationperformance of the suspension in the low frequency andlow-frequency resonance areas is similar to that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withactive suspension the vibration isolation performance of thesuspension in the low frequency high frequency and res-onance regions is slightly better
e dynamic response of the sprung mass acceleration ofthe hybrid suspension in semiactive mode under randomroad is shown in Figure 26 e power spectrum of thesprung mass acceleration of the hybrid suspension insemiactive mode is shown in Figure 27
Table 7 shows the sprung mass acceleration RMS testvalues of the hybrid suspension in semiactive mode
From Table 7 compared with passive suspension aw ofthe hybrid suspension in semiactive mode is reduced by1642 Compared with semiactive suspension aw of thehybrid suspension in active mode is reduced by 307 Andfrom Figure 27 when the hybrid suspension is semiactivelycontrolled compared with passive suspension the vibration
Table 3 e dynamic response root mean square values of thehybrid suspension in urban circulation unit
Figure 21 e system energy of the hybrid suspension in urbancirculation units
0 100 200 300 400ndash20000
ndash15000
ndash10000
ndash5000
0
5000
Time (s)
Susp
ensio
n sy
stem
ener
gy (J
)
Active suspensionHybrid suspension
Figure 22e system energy of the hybrid suspension in suburbancirculation units
Table 5 e system energy values of the hybrid suspension (J)
Cycle unit Active suspension Hybrid suspensionUrban minus2165 38Suburban minus15071 minus122
Figure 23 e test system of the hybrid active suspension
Shock and Vibration 13
isolation performance of the suspension in the low frequencyand low-frequency resonance areas is worse than that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withsemiactive suspension the vibration isolation performance ofthe suspension in the low frequency high frequency andresonance regions is slightly better
e energy consumption power and energy-regenerativepower of the hybrid suspension are shown in Figure 28
From Figure 28 in the active mode the average powerconsumption of the ball screw actuator is 2561W In thesemiactive mode the average regenerative power of the ballscrew actuator is 2696W and the average of theMR damperenergy consumption power is 178W erefore the pureaverage power consumption of the hybrid suspension isminus043W and the test results are basically consistent with thesimulation
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spr
ung
mas
s acc
eler
atio
n (m
s2 )
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 24 e dynamic response of the sprung mass accelerationof the hybrid suspension in active mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 25 e power spectrum of the sprung mass acceleration ofthe hybrid suspension in active mode
Table 6 e sprung mass acceleration RMS of the hybrid sus-pension in active mode
Indicators Passivesuspension
Activesuspension
Hybridsuspension
aw (ms2) 13378 09423 08074
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spru
ng m
ass a
ccel
erat
ion
(ms
2 )
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 26 e dynamic responses of the sprung mass accelerationof the hybrid suspension in semiactive mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 27e power spectrums of the sprungmass acceleration ofthe hybrid suspension in semiactive mode
Table 7 e sprung mass acceleration RMS of the hybrid sus-pension in semiactive mode
Indicators Passivesuspension
Semiactivesuspension
Hybridsuspension
aw (ms2) 13378 11536 11182
14 Shock and Vibration
6 Conclusion
(1) A ball screw actuator andMR damper are introducedinto vehicle suspension system and a new kind ofhybrid active suspension structure is put forwarde ball screw actuator is prototyped and thefunctional relationship between the back-EMF co-ecient the electromagnetic torque coecient of themotor and the suspension vibration speed is ob-tained by test analyses And the active output me-chanical properties of the ball screw actuator aretested and the results show that the actuator hasgood active output force characteristics
(2) e inshyuences of the variable damping value of thesuspension system on the riding comfort handlingand stability and energy consumption characteristicsof the hybrid suspension in the active control modeare analyzeden the optimal damping values of thehybrid suspension at dierent vehicle speeds aredesigned e eects of electromagnetic dampingforce on the actual semiactive force and the systemenergy of the suspension in the semiactive controlmode are analyzed and then the hybrid suspensionwith semiactive mode which has electromagneticdamping force feedback adjustment is designed Onthis basis a multimode coordinated control strategyfor the hybrid suspension is designed
(3) e damping performance and energy consumptioncharacteristics of the hybrid suspension under cyclicdriving condition are simulated by MATLABSimulink software and the results show that whenin the urban circulation unit compared with activesuspension aw and DTLrms of the hybrid suspensionare reduced by 1322 and 840 respectively Andwhen in the suburban circulation unit comparedwith the active suspension aw and DTLrms of thehybrid suspension are reduced by 1199 and 732respectively e pure energy of the active suspen-sion system during the entire operation cycle is
minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid sus-pension system basically realizes energy self-poweredin theory
(4) e eectiveness verication test of the hybrid sus-pension in active mode and semiactive mode controlis carried out and the results show that when thehybrid suspension is actively controlled comparedwith active suspension aw of the hybrid suspension inactive mode is reduced by 1432 When the hybridsuspension is semiactively controlled compared withsemiactive suspension aw of the hybrid suspension inactive mode is reduced by 307 e test and sim-ulation results are basically consistent and the testveries the correctness of the simulation
Data Availability
e data used to support the ndings of this study are in-cluded within the supplementary information les And thedata used to support the ndings of this study are availablefrom the corresponding author upon request
Conflicts of Interest
e authors declare that there are no conshyicts of interestregarding the publication of this paper
Acknowledgments
is work was supported by the National Natural ScienceFoundation of China (Grant no 51775426) Service LocalSpecial Program Support Project of Shaanxi ProvincialEducation Department (Grant no 17JF017) and XirsquoanScience and Technology Program Funding Project (Grantno 2017079CGRC042-XAKD007)
Supplementary Materials
e supplementary materials are this articlersquos experimentaldata including (1) MR damperrsquos characteristic test data (2)the motor counter electromotive forcersquos peak test data (3)the ball screw actuatorrsquos active output force test data (4) thehybrid suspensionrsquos sprung mass acceleration test data inactive mode (5) the hybrid suspensionrsquos sprung mass ac-celeration test data in semiactive mode and (6) the hybridsuspensionrsquos energy consumption power and energy re-generation power test data (Supplementary Materials)
References
[1] M Montazeri-Gh and O Kavianipour ldquoInvestigation of theactive electromagnetic suspension system considering hybridcontrol strategyrdquo Proceedings of the Institution of MechanicalEngineers Part C Journal of Mechanical Engineering Sciencevol 228 no 10 pp 1658ndash1669 2013
[2] H E Tseng and D Hrovat ldquoState of the art survey active andsemi-active suspension controlrdquo Vehicle System Dynamicsvol 53 no 7 pp 1034ndash1062 2015
0 1 2 3 4 50
20
40
60
80
Time (s)
Pow
er (W
)
Energy consumption power in active modeEnergy regenerative power in semiactive modeEnergy consumption power in semiactive mode
Figure 28 e energy consumption power and energy-re-generative power of the hybrid suspension
Shock and Vibration 15
[3] F Kou J Du Z Wang D Li and J Xu ldquoNonlinear modelingand coordinate optimization of a semi-active energy re-generative suspension with an electro-hydraulic actuatorrdquoChina Mechanical Engineering vol 28 no 14 pp 1701ndash17072017
[4] S H Zareh A Sarrafan A A A Khayyat and A ZabihollahldquoIntelligent semi-active vibration control of eleven degrees offreedom suspension system using magnetorheologicaldampersrdquo Journal of Mechanical Science and Technologyvol 26 no 3 pp 323ndash334 2012
[5] S A Chen X Li L J Zhao Y X Wang and Y B KimldquoDevelopment of a control method for an electromagneticsemi-active suspension reclaiming energy with varying chargevoltage in stepsrdquo International Journal of Automotive Tech-nology vol 16 no 5 pp 765ndash773 2015
[6] K Nakano Y Suda and S Nakadai ldquoSelf-powered activevibration control using a single electric actuatorrdquo Journal ofSound and Vibration vol 260 no 2 pp 213ndash235 2003
[7] K Nakano ldquoCombined type self-powered active vibrationcontrol of truck cabinsrdquo Vehicle System Dynamics vol 41no 6 pp 449ndash473 2004
[8] K Huang Y C Zhang F Yu and Y H Gu ldquoCoordinateoptimization for synthetical performance of electrical energy-regenerative active suspensionrdquo Journal of Shanghai JiaotongUniversity vol 43 no 2 pp 226ndash230 2009
[9] K Huang F Yu and Y C Zhang ldquoActive control of energy-regenerative electromagnetic suspension based on energy flowanalysisrdquo Journal of Shanghai Jiaotong University vol 45no 67 pp 1068ndash1073 2011
[10] D S Huang J Q Zhang Y L Liu L Yi and X Y WangldquoPerformance of a novel energy-regenerative active suspen-sion systemrdquo Journal of Chongqing University (English Edi-tion) vol 14 no 3 pp 109ndash118 2015
[11] B Ebrahimi H Bolandhemmat M B Khamesee andF Golnaraghi ldquoA hybrid electromagnetic shock absorber foractive vehicle suspension systemsrdquo Vehicle System Dynamicsvol 49 no 1-2 pp 311ndash332 2011
[12] S Tang L Chen R Wang X Sun and D Shi ldquoResearch onoptimal control of active suspension based on damping multi-modal switchingrdquo Journal of Guangxi University (NaturalScience) vol 39 no 2 pp 300ndash307 2014
[13] R Wang X Ma R Ding X Meng and L Chen ldquoResearchof multi-mode switching control system for hybrid sus-pension based on model referencerdquo Transactions of theChinese Society for Agricultural Machinery vol 48 no 7pp 353ndash360 2017
[14] R Wang Y Qian R Ding X Meng and J Xie ldquoDesign andtests for damping-stiffness of a hybrid electromagnetic sus-pension based on LQGrdquo Journal of Vibration and Shockvol 37 no 3 pp 61ndash65 2017
[15] B Vanavil K K Chaitanya and A S Rao ldquoImproved PIDcontroller design for unstable time delay processes based ondirect synthesis method and maximum sensitivityrdquo Taylorand Francis vol 46 no 8 pp 1349ndash1366 2015
[16] I Mihai and F Andronic ldquoBehavior of a semi-active sus-pension system versus a passive suspension system on anuneven road surfacerdquo Mechanics vol 20 no 1 pp 64ndash692014
[17] D Ngoduy ldquoLinear stability of a generalized multi-anticipative car following model with time delaysrdquo Com-munications in Nonlinear Science and Numerical Simulationvol 22 no 1 pp 420ndash426 2015
[18] H Zhang E Wang F Min R Subash and C Su ldquoSkyhook-based semi-active control of full-vehicle suspension with
magneto-rheological dampersrdquo Chinese Journal of Mechan-ical Engineering vol 26 no 3 pp 498ndash505 2013
[19] F Kou ldquoAn experimental study on the dynamic character-istics of vehicle semi-active seat suspension with magneto-rheological damperrdquo Automotive Engineering vol 37 no 11pp 1346ndash1352 2015
[20] Z Feng S Chen and Y Liang ldquoAn experimental study on thedynamic characteristics of a megneto-rheological semi-activesuspensionrdquoAutomotive Engineering vol 35 no 1 pp 72ndash772013
[21] Z Li and L Xu A New Type of Magnetorheological Damperand Semi-Active coNtrol Design-eory Science Press BeijingChina 2012
[22] B L J Gysen J J H Paulides J L G Janssen andE A Lomonova ldquoActive electromagnetic suspension systemfor improved vehicle dynamicsrdquo IEEE Transactions on Ve-hicular Technology vol 59 no 3 pp 1156ndash1163 2010
[23] Q N Wang S S Liu W H Wang and H Wei ldquoStructuredesign and parameter matching of ball-screw regenerativedamperrdquo Journal of Jilin University (Engineering and Tech-nology Edition) vol 42 no 5 pp 1100ndash1106 2012
[24] S Chen R He and S Lu ldquoEvaluating system of reclaimingenergy suspension comprehensive performancerdquo Trans-actions of the Chinese Society for Agricultural Machineryvol 37 no 7 pp 14ndash18 2006
[25] Y Fan Control Research on Vehicle Suspension with Electro-Hydrostatic Actuator Xirsquoan University of Science and Tech-nology Xirsquoan China 2017
[26] Z Yu Automobile -eory China Machine Press BeijingChina 2009
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Navigation and Observation
International Journal of
Hindawi
wwwhindawicom Volume 2018
Advances in
Multimedia
Submit your manuscripts atwwwhindawicom
isolation performance of the suspension in the low frequencyand low-frequency resonance areas is worse than that of thepassive suspension and the vibration isolation performancein the high frequency and high-frequency resonance areas isbetter than that of the passive suspension Compared withsemiactive suspension the vibration isolation performance ofthe suspension in the low frequency high frequency andresonance regions is slightly better
e energy consumption power and energy-regenerativepower of the hybrid suspension are shown in Figure 28
From Figure 28 in the active mode the average powerconsumption of the ball screw actuator is 2561W In thesemiactive mode the average regenerative power of the ballscrew actuator is 2696W and the average of theMR damperenergy consumption power is 178W erefore the pureaverage power consumption of the hybrid suspension isminus043W and the test results are basically consistent with thesimulation
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spr
ung
mas
s acc
eler
atio
n (m
s2 )
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 24 e dynamic response of the sprung mass accelerationof the hybrid suspension in active mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionActive suspensionHybrid suspension in active mode
Figure 25 e power spectrum of the sprung mass acceleration ofthe hybrid suspension in active mode
Table 6 e sprung mass acceleration RMS of the hybrid sus-pension in active mode
Indicators Passivesuspension
Activesuspension
Hybridsuspension
aw (ms2) 13378 09423 08074
0 1 2 3 4 5ndash4
ndash2
0
2
4
6
Time (s)
Spru
ng m
ass a
ccel
erat
ion
(ms
2 )
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 26 e dynamic responses of the sprung mass accelerationof the hybrid suspension in semiactive mode
0 10 20 30 4000
05
10
15
20
Frequency (Hz)
Am
plitu
de (m
middotsndash2)2 middotH
zndash1
Passive suspensionSemiactive suspensionHybrid suspension in semiactive mode
Figure 27e power spectrums of the sprungmass acceleration ofthe hybrid suspension in semiactive mode
Table 7 e sprung mass acceleration RMS of the hybrid sus-pension in semiactive mode
Indicators Passivesuspension
Semiactivesuspension
Hybridsuspension
aw (ms2) 13378 11536 11182
14 Shock and Vibration
6 Conclusion
(1) A ball screw actuator andMR damper are introducedinto vehicle suspension system and a new kind ofhybrid active suspension structure is put forwarde ball screw actuator is prototyped and thefunctional relationship between the back-EMF co-ecient the electromagnetic torque coecient of themotor and the suspension vibration speed is ob-tained by test analyses And the active output me-chanical properties of the ball screw actuator aretested and the results show that the actuator hasgood active output force characteristics
(2) e inshyuences of the variable damping value of thesuspension system on the riding comfort handlingand stability and energy consumption characteristicsof the hybrid suspension in the active control modeare analyzeden the optimal damping values of thehybrid suspension at dierent vehicle speeds aredesigned e eects of electromagnetic dampingforce on the actual semiactive force and the systemenergy of the suspension in the semiactive controlmode are analyzed and then the hybrid suspensionwith semiactive mode which has electromagneticdamping force feedback adjustment is designed Onthis basis a multimode coordinated control strategyfor the hybrid suspension is designed
(3) e damping performance and energy consumptioncharacteristics of the hybrid suspension under cyclicdriving condition are simulated by MATLABSimulink software and the results show that whenin the urban circulation unit compared with activesuspension aw and DTLrms of the hybrid suspensionare reduced by 1322 and 840 respectively Andwhen in the suburban circulation unit comparedwith the active suspension aw and DTLrms of thehybrid suspension are reduced by 1199 and 732respectively e pure energy of the active suspen-sion system during the entire operation cycle is
minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid sus-pension system basically realizes energy self-poweredin theory
(4) e eectiveness verication test of the hybrid sus-pension in active mode and semiactive mode controlis carried out and the results show that when thehybrid suspension is actively controlled comparedwith active suspension aw of the hybrid suspension inactive mode is reduced by 1432 When the hybridsuspension is semiactively controlled compared withsemiactive suspension aw of the hybrid suspension inactive mode is reduced by 307 e test and sim-ulation results are basically consistent and the testveries the correctness of the simulation
Data Availability
e data used to support the ndings of this study are in-cluded within the supplementary information les And thedata used to support the ndings of this study are availablefrom the corresponding author upon request
Conflicts of Interest
e authors declare that there are no conshyicts of interestregarding the publication of this paper
Acknowledgments
is work was supported by the National Natural ScienceFoundation of China (Grant no 51775426) Service LocalSpecial Program Support Project of Shaanxi ProvincialEducation Department (Grant no 17JF017) and XirsquoanScience and Technology Program Funding Project (Grantno 2017079CGRC042-XAKD007)
Supplementary Materials
e supplementary materials are this articlersquos experimentaldata including (1) MR damperrsquos characteristic test data (2)the motor counter electromotive forcersquos peak test data (3)the ball screw actuatorrsquos active output force test data (4) thehybrid suspensionrsquos sprung mass acceleration test data inactive mode (5) the hybrid suspensionrsquos sprung mass ac-celeration test data in semiactive mode and (6) the hybridsuspensionrsquos energy consumption power and energy re-generation power test data (Supplementary Materials)
References
[1] M Montazeri-Gh and O Kavianipour ldquoInvestigation of theactive electromagnetic suspension system considering hybridcontrol strategyrdquo Proceedings of the Institution of MechanicalEngineers Part C Journal of Mechanical Engineering Sciencevol 228 no 10 pp 1658ndash1669 2013
[2] H E Tseng and D Hrovat ldquoState of the art survey active andsemi-active suspension controlrdquo Vehicle System Dynamicsvol 53 no 7 pp 1034ndash1062 2015
0 1 2 3 4 50
20
40
60
80
Time (s)
Pow
er (W
)
Energy consumption power in active modeEnergy regenerative power in semiactive modeEnergy consumption power in semiactive mode
Figure 28 e energy consumption power and energy-re-generative power of the hybrid suspension
Shock and Vibration 15
[3] F Kou J Du Z Wang D Li and J Xu ldquoNonlinear modelingand coordinate optimization of a semi-active energy re-generative suspension with an electro-hydraulic actuatorrdquoChina Mechanical Engineering vol 28 no 14 pp 1701ndash17072017
[4] S H Zareh A Sarrafan A A A Khayyat and A ZabihollahldquoIntelligent semi-active vibration control of eleven degrees offreedom suspension system using magnetorheologicaldampersrdquo Journal of Mechanical Science and Technologyvol 26 no 3 pp 323ndash334 2012
[5] S A Chen X Li L J Zhao Y X Wang and Y B KimldquoDevelopment of a control method for an electromagneticsemi-active suspension reclaiming energy with varying chargevoltage in stepsrdquo International Journal of Automotive Tech-nology vol 16 no 5 pp 765ndash773 2015
[6] K Nakano Y Suda and S Nakadai ldquoSelf-powered activevibration control using a single electric actuatorrdquo Journal ofSound and Vibration vol 260 no 2 pp 213ndash235 2003
[7] K Nakano ldquoCombined type self-powered active vibrationcontrol of truck cabinsrdquo Vehicle System Dynamics vol 41no 6 pp 449ndash473 2004
[8] K Huang Y C Zhang F Yu and Y H Gu ldquoCoordinateoptimization for synthetical performance of electrical energy-regenerative active suspensionrdquo Journal of Shanghai JiaotongUniversity vol 43 no 2 pp 226ndash230 2009
[9] K Huang F Yu and Y C Zhang ldquoActive control of energy-regenerative electromagnetic suspension based on energy flowanalysisrdquo Journal of Shanghai Jiaotong University vol 45no 67 pp 1068ndash1073 2011
[10] D S Huang J Q Zhang Y L Liu L Yi and X Y WangldquoPerformance of a novel energy-regenerative active suspen-sion systemrdquo Journal of Chongqing University (English Edi-tion) vol 14 no 3 pp 109ndash118 2015
[11] B Ebrahimi H Bolandhemmat M B Khamesee andF Golnaraghi ldquoA hybrid electromagnetic shock absorber foractive vehicle suspension systemsrdquo Vehicle System Dynamicsvol 49 no 1-2 pp 311ndash332 2011
[12] S Tang L Chen R Wang X Sun and D Shi ldquoResearch onoptimal control of active suspension based on damping multi-modal switchingrdquo Journal of Guangxi University (NaturalScience) vol 39 no 2 pp 300ndash307 2014
[13] R Wang X Ma R Ding X Meng and L Chen ldquoResearchof multi-mode switching control system for hybrid sus-pension based on model referencerdquo Transactions of theChinese Society for Agricultural Machinery vol 48 no 7pp 353ndash360 2017
[14] R Wang Y Qian R Ding X Meng and J Xie ldquoDesign andtests for damping-stiffness of a hybrid electromagnetic sus-pension based on LQGrdquo Journal of Vibration and Shockvol 37 no 3 pp 61ndash65 2017
[15] B Vanavil K K Chaitanya and A S Rao ldquoImproved PIDcontroller design for unstable time delay processes based ondirect synthesis method and maximum sensitivityrdquo Taylorand Francis vol 46 no 8 pp 1349ndash1366 2015
[16] I Mihai and F Andronic ldquoBehavior of a semi-active sus-pension system versus a passive suspension system on anuneven road surfacerdquo Mechanics vol 20 no 1 pp 64ndash692014
[17] D Ngoduy ldquoLinear stability of a generalized multi-anticipative car following model with time delaysrdquo Com-munications in Nonlinear Science and Numerical Simulationvol 22 no 1 pp 420ndash426 2015
[18] H Zhang E Wang F Min R Subash and C Su ldquoSkyhook-based semi-active control of full-vehicle suspension with
magneto-rheological dampersrdquo Chinese Journal of Mechan-ical Engineering vol 26 no 3 pp 498ndash505 2013
[19] F Kou ldquoAn experimental study on the dynamic character-istics of vehicle semi-active seat suspension with magneto-rheological damperrdquo Automotive Engineering vol 37 no 11pp 1346ndash1352 2015
[20] Z Feng S Chen and Y Liang ldquoAn experimental study on thedynamic characteristics of a megneto-rheological semi-activesuspensionrdquoAutomotive Engineering vol 35 no 1 pp 72ndash772013
[21] Z Li and L Xu A New Type of Magnetorheological Damperand Semi-Active coNtrol Design-eory Science Press BeijingChina 2012
[22] B L J Gysen J J H Paulides J L G Janssen andE A Lomonova ldquoActive electromagnetic suspension systemfor improved vehicle dynamicsrdquo IEEE Transactions on Ve-hicular Technology vol 59 no 3 pp 1156ndash1163 2010
[23] Q N Wang S S Liu W H Wang and H Wei ldquoStructuredesign and parameter matching of ball-screw regenerativedamperrdquo Journal of Jilin University (Engineering and Tech-nology Edition) vol 42 no 5 pp 1100ndash1106 2012
[24] S Chen R He and S Lu ldquoEvaluating system of reclaimingenergy suspension comprehensive performancerdquo Trans-actions of the Chinese Society for Agricultural Machineryvol 37 no 7 pp 14ndash18 2006
[25] Y Fan Control Research on Vehicle Suspension with Electro-Hydrostatic Actuator Xirsquoan University of Science and Tech-nology Xirsquoan China 2017
[26] Z Yu Automobile -eory China Machine Press BeijingChina 2009
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Navigation and Observation
International Journal of
Hindawi
wwwhindawicom Volume 2018
Advances in
Multimedia
Submit your manuscripts atwwwhindawicom
6 Conclusion
(1) A ball screw actuator andMR damper are introducedinto vehicle suspension system and a new kind ofhybrid active suspension structure is put forwarde ball screw actuator is prototyped and thefunctional relationship between the back-EMF co-ecient the electromagnetic torque coecient of themotor and the suspension vibration speed is ob-tained by test analyses And the active output me-chanical properties of the ball screw actuator aretested and the results show that the actuator hasgood active output force characteristics
(2) e inshyuences of the variable damping value of thesuspension system on the riding comfort handlingand stability and energy consumption characteristicsof the hybrid suspension in the active control modeare analyzeden the optimal damping values of thehybrid suspension at dierent vehicle speeds aredesigned e eects of electromagnetic dampingforce on the actual semiactive force and the systemenergy of the suspension in the semiactive controlmode are analyzed and then the hybrid suspensionwith semiactive mode which has electromagneticdamping force feedback adjustment is designed Onthis basis a multimode coordinated control strategyfor the hybrid suspension is designed
(3) e damping performance and energy consumptioncharacteristics of the hybrid suspension under cyclicdriving condition are simulated by MATLABSimulink software and the results show that whenin the urban circulation unit compared with activesuspension aw and DTLrms of the hybrid suspensionare reduced by 1322 and 840 respectively Andwhen in the suburban circulation unit comparedwith the active suspension aw and DTLrms of thehybrid suspension are reduced by 1199 and 732respectively e pure energy of the active suspen-sion system during the entire operation cycle is
minus23731 J while the pure energy of the hybrid sus-pension system is 38 J erefore the hybrid sus-pension system basically realizes energy self-poweredin theory
(4) e eectiveness verication test of the hybrid sus-pension in active mode and semiactive mode controlis carried out and the results show that when thehybrid suspension is actively controlled comparedwith active suspension aw of the hybrid suspension inactive mode is reduced by 1432 When the hybridsuspension is semiactively controlled compared withsemiactive suspension aw of the hybrid suspension inactive mode is reduced by 307 e test and sim-ulation results are basically consistent and the testveries the correctness of the simulation
Data Availability
e data used to support the ndings of this study are in-cluded within the supplementary information les And thedata used to support the ndings of this study are availablefrom the corresponding author upon request
Conflicts of Interest
e authors declare that there are no conshyicts of interestregarding the publication of this paper
Acknowledgments
is work was supported by the National Natural ScienceFoundation of China (Grant no 51775426) Service LocalSpecial Program Support Project of Shaanxi ProvincialEducation Department (Grant no 17JF017) and XirsquoanScience and Technology Program Funding Project (Grantno 2017079CGRC042-XAKD007)
Supplementary Materials
e supplementary materials are this articlersquos experimentaldata including (1) MR damperrsquos characteristic test data (2)the motor counter electromotive forcersquos peak test data (3)the ball screw actuatorrsquos active output force test data (4) thehybrid suspensionrsquos sprung mass acceleration test data inactive mode (5) the hybrid suspensionrsquos sprung mass ac-celeration test data in semiactive mode and (6) the hybridsuspensionrsquos energy consumption power and energy re-generation power test data (Supplementary Materials)
References
[1] M Montazeri-Gh and O Kavianipour ldquoInvestigation of theactive electromagnetic suspension system considering hybridcontrol strategyrdquo Proceedings of the Institution of MechanicalEngineers Part C Journal of Mechanical Engineering Sciencevol 228 no 10 pp 1658ndash1669 2013
[2] H E Tseng and D Hrovat ldquoState of the art survey active andsemi-active suspension controlrdquo Vehicle System Dynamicsvol 53 no 7 pp 1034ndash1062 2015
0 1 2 3 4 50
20
40
60
80
Time (s)
Pow
er (W
)
Energy consumption power in active modeEnergy regenerative power in semiactive modeEnergy consumption power in semiactive mode
Figure 28 e energy consumption power and energy-re-generative power of the hybrid suspension
Shock and Vibration 15
[3] F Kou J Du Z Wang D Li and J Xu ldquoNonlinear modelingand coordinate optimization of a semi-active energy re-generative suspension with an electro-hydraulic actuatorrdquoChina Mechanical Engineering vol 28 no 14 pp 1701ndash17072017
[4] S H Zareh A Sarrafan A A A Khayyat and A ZabihollahldquoIntelligent semi-active vibration control of eleven degrees offreedom suspension system using magnetorheologicaldampersrdquo Journal of Mechanical Science and Technologyvol 26 no 3 pp 323ndash334 2012
[5] S A Chen X Li L J Zhao Y X Wang and Y B KimldquoDevelopment of a control method for an electromagneticsemi-active suspension reclaiming energy with varying chargevoltage in stepsrdquo International Journal of Automotive Tech-nology vol 16 no 5 pp 765ndash773 2015
[6] K Nakano Y Suda and S Nakadai ldquoSelf-powered activevibration control using a single electric actuatorrdquo Journal ofSound and Vibration vol 260 no 2 pp 213ndash235 2003
[7] K Nakano ldquoCombined type self-powered active vibrationcontrol of truck cabinsrdquo Vehicle System Dynamics vol 41no 6 pp 449ndash473 2004
[8] K Huang Y C Zhang F Yu and Y H Gu ldquoCoordinateoptimization for synthetical performance of electrical energy-regenerative active suspensionrdquo Journal of Shanghai JiaotongUniversity vol 43 no 2 pp 226ndash230 2009
[9] K Huang F Yu and Y C Zhang ldquoActive control of energy-regenerative electromagnetic suspension based on energy flowanalysisrdquo Journal of Shanghai Jiaotong University vol 45no 67 pp 1068ndash1073 2011
[10] D S Huang J Q Zhang Y L Liu L Yi and X Y WangldquoPerformance of a novel energy-regenerative active suspen-sion systemrdquo Journal of Chongqing University (English Edi-tion) vol 14 no 3 pp 109ndash118 2015
[11] B Ebrahimi H Bolandhemmat M B Khamesee andF Golnaraghi ldquoA hybrid electromagnetic shock absorber foractive vehicle suspension systemsrdquo Vehicle System Dynamicsvol 49 no 1-2 pp 311ndash332 2011
[12] S Tang L Chen R Wang X Sun and D Shi ldquoResearch onoptimal control of active suspension based on damping multi-modal switchingrdquo Journal of Guangxi University (NaturalScience) vol 39 no 2 pp 300ndash307 2014
[13] R Wang X Ma R Ding X Meng and L Chen ldquoResearchof multi-mode switching control system for hybrid sus-pension based on model referencerdquo Transactions of theChinese Society for Agricultural Machinery vol 48 no 7pp 353ndash360 2017
[14] R Wang Y Qian R Ding X Meng and J Xie ldquoDesign andtests for damping-stiffness of a hybrid electromagnetic sus-pension based on LQGrdquo Journal of Vibration and Shockvol 37 no 3 pp 61ndash65 2017
[15] B Vanavil K K Chaitanya and A S Rao ldquoImproved PIDcontroller design for unstable time delay processes based ondirect synthesis method and maximum sensitivityrdquo Taylorand Francis vol 46 no 8 pp 1349ndash1366 2015
[16] I Mihai and F Andronic ldquoBehavior of a semi-active sus-pension system versus a passive suspension system on anuneven road surfacerdquo Mechanics vol 20 no 1 pp 64ndash692014
[17] D Ngoduy ldquoLinear stability of a generalized multi-anticipative car following model with time delaysrdquo Com-munications in Nonlinear Science and Numerical Simulationvol 22 no 1 pp 420ndash426 2015
[18] H Zhang E Wang F Min R Subash and C Su ldquoSkyhook-based semi-active control of full-vehicle suspension with
magneto-rheological dampersrdquo Chinese Journal of Mechan-ical Engineering vol 26 no 3 pp 498ndash505 2013
[19] F Kou ldquoAn experimental study on the dynamic character-istics of vehicle semi-active seat suspension with magneto-rheological damperrdquo Automotive Engineering vol 37 no 11pp 1346ndash1352 2015
[20] Z Feng S Chen and Y Liang ldquoAn experimental study on thedynamic characteristics of a megneto-rheological semi-activesuspensionrdquoAutomotive Engineering vol 35 no 1 pp 72ndash772013
[21] Z Li and L Xu A New Type of Magnetorheological Damperand Semi-Active coNtrol Design-eory Science Press BeijingChina 2012
[22] B L J Gysen J J H Paulides J L G Janssen andE A Lomonova ldquoActive electromagnetic suspension systemfor improved vehicle dynamicsrdquo IEEE Transactions on Ve-hicular Technology vol 59 no 3 pp 1156ndash1163 2010
[23] Q N Wang S S Liu W H Wang and H Wei ldquoStructuredesign and parameter matching of ball-screw regenerativedamperrdquo Journal of Jilin University (Engineering and Tech-nology Edition) vol 42 no 5 pp 1100ndash1106 2012
[24] S Chen R He and S Lu ldquoEvaluating system of reclaimingenergy suspension comprehensive performancerdquo Trans-actions of the Chinese Society for Agricultural Machineryvol 37 no 7 pp 14ndash18 2006
[25] Y Fan Control Research on Vehicle Suspension with Electro-Hydrostatic Actuator Xirsquoan University of Science and Tech-nology Xirsquoan China 2017
[26] Z Yu Automobile -eory China Machine Press BeijingChina 2009
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Navigation and Observation
International Journal of
Hindawi
wwwhindawicom Volume 2018
Advances in
Multimedia
Submit your manuscripts atwwwhindawicom
[3] F Kou J Du Z Wang D Li and J Xu ldquoNonlinear modelingand coordinate optimization of a semi-active energy re-generative suspension with an electro-hydraulic actuatorrdquoChina Mechanical Engineering vol 28 no 14 pp 1701ndash17072017
[4] S H Zareh A Sarrafan A A A Khayyat and A ZabihollahldquoIntelligent semi-active vibration control of eleven degrees offreedom suspension system using magnetorheologicaldampersrdquo Journal of Mechanical Science and Technologyvol 26 no 3 pp 323ndash334 2012
[5] S A Chen X Li L J Zhao Y X Wang and Y B KimldquoDevelopment of a control method for an electromagneticsemi-active suspension reclaiming energy with varying chargevoltage in stepsrdquo International Journal of Automotive Tech-nology vol 16 no 5 pp 765ndash773 2015
[6] K Nakano Y Suda and S Nakadai ldquoSelf-powered activevibration control using a single electric actuatorrdquo Journal ofSound and Vibration vol 260 no 2 pp 213ndash235 2003
[7] K Nakano ldquoCombined type self-powered active vibrationcontrol of truck cabinsrdquo Vehicle System Dynamics vol 41no 6 pp 449ndash473 2004
[8] K Huang Y C Zhang F Yu and Y H Gu ldquoCoordinateoptimization for synthetical performance of electrical energy-regenerative active suspensionrdquo Journal of Shanghai JiaotongUniversity vol 43 no 2 pp 226ndash230 2009
[9] K Huang F Yu and Y C Zhang ldquoActive control of energy-regenerative electromagnetic suspension based on energy flowanalysisrdquo Journal of Shanghai Jiaotong University vol 45no 67 pp 1068ndash1073 2011
[10] D S Huang J Q Zhang Y L Liu L Yi and X Y WangldquoPerformance of a novel energy-regenerative active suspen-sion systemrdquo Journal of Chongqing University (English Edi-tion) vol 14 no 3 pp 109ndash118 2015
[11] B Ebrahimi H Bolandhemmat M B Khamesee andF Golnaraghi ldquoA hybrid electromagnetic shock absorber foractive vehicle suspension systemsrdquo Vehicle System Dynamicsvol 49 no 1-2 pp 311ndash332 2011
[12] S Tang L Chen R Wang X Sun and D Shi ldquoResearch onoptimal control of active suspension based on damping multi-modal switchingrdquo Journal of Guangxi University (NaturalScience) vol 39 no 2 pp 300ndash307 2014
[13] R Wang X Ma R Ding X Meng and L Chen ldquoResearchof multi-mode switching control system for hybrid sus-pension based on model referencerdquo Transactions of theChinese Society for Agricultural Machinery vol 48 no 7pp 353ndash360 2017
[14] R Wang Y Qian R Ding X Meng and J Xie ldquoDesign andtests for damping-stiffness of a hybrid electromagnetic sus-pension based on LQGrdquo Journal of Vibration and Shockvol 37 no 3 pp 61ndash65 2017
[15] B Vanavil K K Chaitanya and A S Rao ldquoImproved PIDcontroller design for unstable time delay processes based ondirect synthesis method and maximum sensitivityrdquo Taylorand Francis vol 46 no 8 pp 1349ndash1366 2015
[16] I Mihai and F Andronic ldquoBehavior of a semi-active sus-pension system versus a passive suspension system on anuneven road surfacerdquo Mechanics vol 20 no 1 pp 64ndash692014
[17] D Ngoduy ldquoLinear stability of a generalized multi-anticipative car following model with time delaysrdquo Com-munications in Nonlinear Science and Numerical Simulationvol 22 no 1 pp 420ndash426 2015
[18] H Zhang E Wang F Min R Subash and C Su ldquoSkyhook-based semi-active control of full-vehicle suspension with
magneto-rheological dampersrdquo Chinese Journal of Mechan-ical Engineering vol 26 no 3 pp 498ndash505 2013
[19] F Kou ldquoAn experimental study on the dynamic character-istics of vehicle semi-active seat suspension with magneto-rheological damperrdquo Automotive Engineering vol 37 no 11pp 1346ndash1352 2015
[20] Z Feng S Chen and Y Liang ldquoAn experimental study on thedynamic characteristics of a megneto-rheological semi-activesuspensionrdquoAutomotive Engineering vol 35 no 1 pp 72ndash772013
[21] Z Li and L Xu A New Type of Magnetorheological Damperand Semi-Active coNtrol Design-eory Science Press BeijingChina 2012
[22] B L J Gysen J J H Paulides J L G Janssen andE A Lomonova ldquoActive electromagnetic suspension systemfor improved vehicle dynamicsrdquo IEEE Transactions on Ve-hicular Technology vol 59 no 3 pp 1156ndash1163 2010
[23] Q N Wang S S Liu W H Wang and H Wei ldquoStructuredesign and parameter matching of ball-screw regenerativedamperrdquo Journal of Jilin University (Engineering and Tech-nology Edition) vol 42 no 5 pp 1100ndash1106 2012
[24] S Chen R He and S Lu ldquoEvaluating system of reclaimingenergy suspension comprehensive performancerdquo Trans-actions of the Chinese Society for Agricultural Machineryvol 37 no 7 pp 14ndash18 2006
[25] Y Fan Control Research on Vehicle Suspension with Electro-Hydrostatic Actuator Xirsquoan University of Science and Tech-nology Xirsquoan China 2017
[26] Z Yu Automobile -eory China Machine Press BeijingChina 2009
![ResearchArticle MulticlassEventClassificationfromText2020/11/03 · script. Its grammatical structure is different from other languages. (1)Subject-object-verb(SOV)sentencestructure[14]](https://static.documents.pub/doc/80x56/610cf929c281e47eb17c0612/researcharticle-multiclasseventclassificationfromtext-20201103-script-its.jpg)
