30.5.2012
1
SIMA – Raw Data Simulation Software for the Development and Validation of Algorithms for GNSS and MEMS based Multi‐Sensor Navigation
Platforms
30.05.2012 Seite 1
Andreas Hoscislawski
HS‐Karlsruhe, Germany
FIG working week 2012 ‐ Rome
NAVIGATION STATE & FRAMES
30.05.2012 Seite 2
t B, L, h | v , v , v | r, p, yNavigation state vector:
xn
yn
zn
xb
yb
zb
y
zb
yb
xb
Body
position (B,L,h) + velocity ( , , ) + orientation (r,p,y)
30.5.2012
2
SENSORS FOR ROBUST AND GLOBAL APPLICATIONS
30.05.2012
References: Inertial Space and Gravity Field
1.) GNSS
2.) Accelerometers
3.) Gyroscopes
4.) Magnetic field sensors
5.) Inclinometers
References: Inertial Space
References: Earth Magnetic Field
References: Gravity Field
References: Inertial Space or e‐frame
Seite 3
SENSOR CONCEPT FOR STATE ESTIMATION
30.05.2012 Seite 4
General concept for robust algorithms & sensor simulation:
„Multiplatform‐“(several platforms (p) navigate one body (b))
and„Multisensor‐Leverarm‐“ – Concept
(several coordinated sensors on each platform)
30.5.2012
3
SIMA: SIMULATION OF MULTISENSOR ARRAYS
• Numerical comparison of optimized sensor platforms
• Numerical proof of functionality of new platforms
– with redundant sensors
– with sensors in motion
• Further system tests:
– Can additional parameters be estimated?
– Filter reaction on gross errors?
– Filter reactions on different trajectories?
• Simplified implementation because true numerical values are known
• Reference state known from trajectory model
30.05.2012 Seite 5
SIMA LEVER ARM CONCEPT & PARAMETRIZATION
30.05.2012 Seite 6
Xe
Ye
Ze
Yp
Xp
Zp
Ys
Xs
Zs
t(t) esensor
sensitive axisof sensor j
platform i
body
x(t)ebody
x(t)esensor
t eplat
t t t,
5 LA‐parameter sensor j
, α, δ,
6 LA‐parameter platform i
, , , ε , ε , ε
30.5.2012
4
SIMA LEVER ARM CONCEPT & M‐FRAME
Xe
Ye
Ze
Yp
Xp
Zp
Ys
Xs
Zs
t(t) ssensor
sensitive axisof sensor j
platform i
body
x(t)mbody
x(t)esensor
t pplat
t _ t t t,
Xm
Zmx
eorigin m‐frame
Seite 7
tt t
t
,
.
LEVER ARM CONCEPT – SENSOR VELOCITY & ACCELERATION
Seite 8
tt t
t
,
constant
constant
, , , , , , , , , , , , ,
Necessary parameters for observation modeling:
30.5.2012
5
TRAJECTORY GENERATION
• Trajectory parameters:
• Standard models: straight line, circle, helix, in rest, rotating, 2D‐trajectory
• Example: Body orientation in a circle
30.05.2012 Seite 9
t t t t t t
y t atanv
v
v =cos(v t / R)
v sin vt/R v
GNSS OBSERVATIONS
• GNSS position
• GNSS velocity
30.05.2012 Seite 10
t t t,
, ,
t t
t,
, ,
constant I
constant
30.5.2012
6
l , , , , , c Δt , Δt , λ N λ D,
ΔIon ΔTrop
GNSS OBSERVATIONS
• Raw data observation equations:
– pseudorange
– phase
– Doppler
30.05.2012 Seite 11
, , , , c Δt , Δt , ,ΔIon ΔTrop
l , f 1 ,
cf
,
sX
sY
sZ
ACCELEROMETER OBSERVATIONS
• Navigation equation in the inertial frame:
• Navigation equation in the earth frame:
• Rotation to the s‐frame:
• Adding sensor errors:
30.05.2012 Seite 12
t2 .
a , 1 0 0 ∙
l , a , ∙ κ b n .
t
30.5.2012
7
• Gyro observation model:
• for one sensor j on platform i:
• Adding sensor errors:
GYROSCOPE OBSERVATIONS
30.05.2012 Seite 13
.
.
ω , 1 0 0 ∙ .
l,
ω , ∙ κ b n .
ωsis, y
ωsis, z
ωsis, x
MEMS3D‐Gyroskop22 x 22 mm
MAGNETIC FIELD OBSERVATIONS
• Magnetic field observation:
• World Magnetic Model 2010 from NOAA (National Oceanic andAtmospheric Administration) & BGS (British Geological Survey):
• for one sensor j on platform i:
• Error model:
30.05.2012 Seite 14
, t
m,
1 0 0 ∙
l,
m,
n .
30.5.2012
8
• Observation equation inclinometer:
• Rotation of sensitive axis in s‐frame to LAV:
• Adding sensor errors:
INCLINOMETER OBSERVATIONS
30.05.2012 Seite 15
θ , cos
0 0 1
1 0 0
,
l θ c n , .
direction of gravity:
SIMA – GUI
30.05.2012 Seite 16
30.5.2012
9
EXAMPLE – ATTITUDE HEADING REFERENCE SYSTEM
• Navigation state vector
• Trajectory: Body rotates in rest
• Accelerometer biases:
30.05.2012 Seite 17
t , ,
1.0 2.0 3.0
0,0
10.0 20.0 30.0
EXAMPLE – ATTITUDE HEADING REFERENCE SYSTEM
30.05.2012 Seite 18
Kalmal filtered pitch angle: Kalmal filtered accelerometer biases
30.5.2012
10
CONCLUSION
• SIMAs features:
– arbitrary number of different types of sensors
– freely open platform design
– consideration of the lever‐arm effects
– modeling of sensorerrors
– different trajectories
– known reference data for filter validation
• Perspective:
– enhanced error modeling
– adding additional trajectories
– …
30.05.2012 Seite 19
Thank you for your attention!
www.navka.de
30.05.2012 Seite 20
SIMA available at:
30.5.2012
11
ENHANCED NAVIGATION‐ALGORITHMS
• Platform optimization in a similar manner as in the conventionalclassification in the optimization of geodetical nets:
– design of 0th order: choise of the appropriate sensor type
– design of 1th order: choice of optimal sensor position and orientationon platform at given variance for the observationsand system state
– design of 2nd order: choice of optimal observation accuracy at givenplatform design and variance of the system state
– design of 3rd order: choice of additional sensors to optimize givenplatform design
30.05.2012 Seite 21
Sensor raw data simulation tool required
• Sensor design differs in
– sensor type
– sensor quantity
– sensor quality
– location
• Different sensor designs for different applications depends on:
– navigation parameters
– body tractory
– required accuracy
MULTI‐SENSOR‐ALGORITHMS DEVELOPMENT
30.05.2012 Seite 22
Sensor raw data simulation tool required