+ All Categories
Home > Documents > Vehicle Counting in Heterogeneous Traffic Condition

Vehicle Counting in Heterogeneous Traffic Condition

Date post: 14-Apr-2018
Category:
Upload: journalofict
View: 221 times
Download: 0 times
Share this document with a friend

of 5

Transcript
  • 7/29/2019 Vehicle Counting in Heterogeneous Traffic Condition

    1/5

  • 7/29/2019 Vehicle Counting in Heterogeneous Traffic Condition

    2/5

  • 7/29/2019 Vehicle Counting in Heterogeneous Traffic Condition

    3/5

    20

    32Kbytes of programmable memory; it also supportsInter-Integrated circuits (I2C) protocol and has 23Programmable I/O lines. The board has Real timeclock (DS 1307), which keeps track of Date(dd/mm/yy) and time (hh/mm/ss). It also providesIsolation between Microcontroller and LD 400

    [10

    ],[11] with ULN 2803(NPN Darlington pair) fromhigh surge current and AC voltage. Microcontrollerhas been interfaced with a computer using serialcommunication and seven segment display or LCDcan be interfaced for displaying output.

    6 VEHICLE COUNTING:

    We have analyzed theoretically many of the sensor sys-tem for counting traffic in India. Most of them havedrawback e.g. counting using video camera need morerobust Algorithm and Hardware, problem like occlusion,vehicle shadowing, a separate module of Algorithm is

    required for different environmental conditions duringrain and snow. Counting Vehicles using radars face dis-tortion problem in rainy season,Preference for sensor technology to be used for detectingtraffic can be made on different criteria such as (i) ease ofinstallation, (ii) ease of Maintenance, (iii) sensor technolo-gy, (iii) Application [2]A combination of these points can be asserted to drawconclusion for selection of sensor technology. Magnetom-eter is more preferred if above points are considered. In-ductive loop on the other hand are less preferred becauseof less ease of installation due to road cutting, wire get-ting damage due to various reason (like foreign solvent inslot of wire, insulation wear and tear, etc). Inductive Loopsensor technology is one of the less expensive technolo-gies for counting of vehicles. Its output is robust in differ-ent types of environmental condition. While cutting roadfor laying wires, corners were cut diagonal instead of 90degrees, diagonal cuts eases stress on wires.

    7 COUNTING ALGORITHM:

    We have installed 8 loops on JVLR in Mumbai. Illustra-tion of system working can be done with sensing 4 loops.Figure 3 gives the dimension of loop installed. Loopwidth being 1.1m and spacing between adjacent loops is

    0.5m at any given time no vehicle will occupy more than3 loops. Below are illustrations of some cases which wehave encountered statistically.These cases have been im-plementated in Algorithm for countingvehicles.

    Figure 3Loop Layout diagram.Case 1: In some cases the vehicle occupies 2 loops, but ittriggers 3 loops. A vehicle being a ferromagnetic materialand due to loop sensitivity, loops get triggered if the ve-hicle passes by its close vicinity. Consequently a vehicle

    going over 2 loops physically, triggers 3 loops, the loopdetector gives output on 3 channels. It has been shown inFigure 4(a), vehicle passing over loops, the dotted linerepresents that though vehicle does not pass physicallyover loop 2, and it is still being detected by loop 2. Figure4 (b) shows the loop detector output which we representit as a timing diagram. Analyzing loop detector output as

    timing diagram eases different cases under consideration.Further in this paper we will deal or represent differentcases with help of timing diagram.

    Figure 4.Case 1Timing DiagramIn Figure 4(a) L1, L2, L3, L4 represent Loop numbers,

    Figure 4(b) represent timing diagram, showing triggeringtime for each loop the variables A[.] and B[.] represent theon-time and off-time for each loop. The amplitude foreach pulse can be varied and we have taken them to be as5V (logic-HIGH). There are two states that the loop detec-tor output can take, digitally HIGH (5V) or LOW (0V).In Case 1 situation, loop detector has received three trig-gering and the count given is as one if;A[3] A[4]

  • 7/29/2019 Vehicle Counting in Heterogeneous Traffic Condition

    4/5

  • 7/29/2019 Vehicle Counting in Heterogeneous Traffic Condition

    5/5

    22

    hicles over adjacent loop at different time of the day. Theaccuracy of classification can be further increased bymeasuring the amplitude of the incoming signal. Eachclass of vehicle has different shape and size with differentferromagnetic content. A two wheeler will cause less var-iation in inductance as compared to a Car or truck. Hence

    the magnitude of signal received from Car or truck willbe higher, this parameter (Magnitude) of signature can beused for classification.

    This system could be useful in controlling traffic inadaptive manner by giving the system output to trafficcontroller. The data generated by the system can be pro-cessed and can be used to divert emergency vehicles ortraffic to a less dense route. The sensor technology is ro-

    bust to environmental conditions and is cheaper in termsof cost to other sensor technologies.

    10ACKNOWLEDGEMENT

    Our sincere gratitude to Department of InformationTechnology for sponsoring the project (09DIT004). A spe-cial thanks to all the people related to the project, as theyhave helped in making the project a partial success. Aspecialthanks to Prof. Gopal Patil for his invaluable helpand my colleagues Ashish Cherian, Ashutosh Bajpai andMohamed Athiq. Lastly I would thank my parents andmy sister Dr. Nivika Gupta for their constant supportthroughout the project.

    11REFERENCES[1] C. Sun(Mar. 2000), An investigation in the use of inductive

    loop signaturesfor vehicle classification, Inst. Transp. Stud.,Univ. California, Berkeley, CA, California PATH Res. Rep.,

    UCB-ITS-PRR-2000.

    [2] J. Gajda, R. Sroka, M. Stencel, and T. Zeqlen (Sep. 1997), Meas-

    urement of road traffic parameters using an inductive single-

    loop detector, in Proc. 9th IMEKO Symp., Glasgow, U.K.

    [3] J.Gajda, R. Sroka, M. Stencel, A. Wajda, and T. Zeqlen (May

    2001), A vehicle classification basedon inductive loop de-tectors, in Proc. 18th IEEE IMTC, vol. 1, pp.460464.

    [4] Seri Oh, Stephen G. Ritchie, and Cheol Oh (Dec. 2001),Real

    time traffic measurement from single loop inductive signa-

    ture, the 81st Annual Meeting of the Transportation Research

    Board, Januray 2002, Washington, D.C.

    [5] Carlos C. Sun, G. S. Arr, S. G. Ritchie (Sep. 2004), Vehicle

    Reidentification Using Multidetector Fusion, IEEE Transac-

    tions on Intelligent Transportation Systems, Vol. 5, No. 3, Sep-

    tember 2004.

    [6] Cheol Oh, S. G. Ritchie (2007),Recognizing vehicle classifica-

    tion information from blade sensor signature, Pattern Recogni-

    tion Letters 28 (2007) 10411049.

    [7] Distribution of population, sex ratio, density and decadal

    growth rate ofpopulation:2011,Office of Registrar General andCensus Commissioner, India.

    [8] Santosh A.Jalihal, Kayitha Ravinder, T.S.Reddy (2005), Traffic

    Characteristics ofIndia,Proceedings of the Eastern Asia Socie-ty for Transportation Studies, Vol. 5, pp. 1009

    1024.

    [9] Lawrence A. Klein, Milton K. Mills and David R.P. Gibson (Oc-

    tober 2006) Publication No.FHWA-HRT-06-108, Traffic Detector

    Handbook: Third Edition, Federal Highway Administration.

    [10] TechnicalManual LD400, Procon Electronics.[11] Technical Manual Loop Profiler, Counter and Accessories

    Ltd.England.

    Swapnil R. Guptadid his B.E from Pune University in 2008, M.Techin Microwave from College of Engineering Pune, in 2010. He is nowcurrently working as Research Assistant in IIT Bombay. His area ofresearch is embedded system, Electronic system design, Artificial

    Intelligence and RF systems.

    Harish K. Pillai did his B.Tech. in Electrical Engineering from IITKharagpur in 1990, M.Tech. in Systems and Control from IIT Bom-

    bay in 1992 and his Ph.D. in Electrical Engineering from IIT Bombay

    in 1997. After spending about 4 years in various universities in Eu-rope, he joined the Department of Electrical Engineering, IIT Bombay

    in 2001, where he is currently a Professor. His current areas of inter-est include Control Theory, Coding Theory, Numerical methods,

    Combinatorial Optimization and Electromagnetic.

    Tom V Mathew is a professor at IIT Bombay. He did his Masters &Doctorate from IIT Madras. He is currently faculty at IIT Bombay. Hisresearch includes Traffic flow modeling, simulation and Transporta-tion network optimization, control and management. Also, his re-

    search includes Economic and financial analysis of transportation

    infrastructure projects , Flexible and rigid pavements analysis & de-sign and Computing techniques like Cellular automata, Genetic Algo-

    rithm, Distributed computing for GA code, Artificial Neural Network,Expert systems and Simulation.


Recommended