Hardware-in-the-loop Simulation of Traction Control Algorithm Based on Fuzzy PID

Hardware-in-the-loop Simulation of Traction Control Algorithm Based on Fuzzy PID

Available online at www.sciencedirect.com Available onlin A ne at www.sciencedirect.co om Enerrgy Proceedia Energy Proce dia Procedia 00 (2011) 00–0...

515KB Sizes 0 Downloads 37 Views

Available online at www.sciencedirect.com Available onlin A ne at www.sciencedirect.co om

Enerrgy Proceedia

Energy Proce dia Procedia 00 (2011) 00–0001685 – 1692 Energy 1600 (2012) www w.elsevier.com/loocate/procedia

20122 Internationnal Conference on Futuure Energy, Environmeent, and Matterials

Haardware--in-the-looop Sim mulation of o Traction Conntrol Algorithhm Baseed on Fu uzzy PID D a Lianng Chua, LiBo L Chaoa,* , Yang Ou O a, WenBoo Lua a

State Key Laboratory L o Automobile Dynamic Simulation, of S Jilin J Universsity, Changchhun, 130025 5, P.R. China *corresponnding authorr

Abstract In this paper, a traction contrrol algorithm baased on fuzzy PID P is proposed d, and a hardwaare Hardware-inn-the-loop (HIL L) w is based on the xPC Target™ producct of MATLAB B® is introducced. The controol algorithm iss test bench which validated usinng the test bench and the testt results of show w that the algo orithm can conttrol the spin off driven wheelss effectively. © 2011 Published by Elsevier B.V. Selection and/or peer-review under responsibility of International Materials Science Society. © 2011 Publiished by Elsevieer Ltd. Selectioon and/or peer-rreview under responsibility of [name organizeer] Keywords:Tracction control; Harrdware-in-the-looop; fuzzy PID;

1.

Introdu uction (Head ding 1)

It is a knoown fact that development of automotive control systems using harrdware-in-the--loop (HIL) iss a very maturre methodologgy. There are several suppliers with available tools likke dSPACE, E ETAS and IPG G, and a great number n of succcessful applications. Besiddes reduced deependence onn physical test systems, HIL L systems havve other virtues: perfectly repeatable tessts and perfecct control of test t environm ment variables. Over the lasst two decadess many autom motive and com mmercial vehiicle makers haave added virrtual testing orr hardware-inn-the-loop (HIL L) testing of electronic e conntrol system to o their repositoory of tools. TCS (Trraction controol system) cann guarantee stability and optimum traaction duringg acceleration, particularly on asymmetriical road surfaaces and durinng cornering maneuvers m by controlling thhe slip ratio off w is typicaally the ratio at which the road r adhesionn coefficient iss driven wheeels to the targeet slip ratio, which at its maxim mum. Typicallly engine torrque control and a driven wh heel braking control are used. In recentt years, the sttudy of TCS focuses f on thee application of modern co ontrol theory to t TCS, TCS for electronicc vehicles andd hybrid vehiccles. In this paaper, a traction control algoorithm based on fuzzy PID D is proposed, and a hardwaare Hardware-in-the-loop (HIL) ( test bennch used to vaalidate the algoorithm is intro oduced. The teest bench whicch is based on n the xPC Tarrget™ producct of MATLA AB® is low cost c and high efficiency. After A the desccription of thee

1876-6102 © 2011 Published by Elsevier B.V. Selection and/or peer-review under responsibility of International Materials Science Society. doi:10.1016/j.egypro.2012.01.261

1686

Liang Chu et al. / Energy Procedia 16 (2012) 1685 – 1692

Author naame / Energy Prrocedia 00 (201 11) 000–000

hardware annd software off test bench annd the control algorithm in detail, d the testt results are lissted in the lastt part of the paper. p 2.

Structu ure of hardw ware-in-the-loop simulation n system The hardw ware in the looop test bench consists of tw wo parts .i.e. th he hardware part p and the sooftware part.

A. Structuree of the test beench The hardware part of the t test bench is based on thhe xPC Targett™ product off MATLAB®, it consists off a target PC, a drive circuiit to drive thee a desktop coomputer as hoost PC, an inddustrial control computer as motor and solenoid s valvees in the Hydrraulic Controll Unit (HCU) and the steppping motor in the electronicc throttle, an electronic e throottle, 4 pressurre sensors, a HCU H and a suit of vehicle hydraulic h brakking system.

Fig. 1. HIL test t bench of traction t controol system During simulation, s thee vehicle model and contrrol algorithmss built in MA ATLAB®, Sim mulink®, and d Stateflow® software are converted to executable code c using thee host PC with Real-Timee Workshop®, W E Embedded Cooder™, Statefflow® Coder™ ™ software, and a a C comppiler. The hostt Real-Time Workshop® PC and targget PC are coommunicatingg through TCP P/IP protocol.. There are tw wo boards in the industriall control com mputer for the purpose p of datta acquisition and PWM geeneration, whicch are PCL 8112PG and PCII 6602 produced by ADV VANTECH annd National Innstruments reespectively. The PCL 812P PG is used to o t analog siggnals from thee pressure senssors. The PCII 6602 is usedd to generated PWM signalss acquisition the to control thhe motor and the t solenoid valves v in the HCU H and the stepping motoor in the electtronic throttle, as it can nott provided enoough power, dedicated d drivee circuit is inttroduced. To control c the braaking pressuree of four brakke cylinders and a to build braking b pressuure actively, a HCU of BO OSCH ESP 8.00 is used. Thee four pressurre sensors which are used to t record the brake pressurre are mounteed near each bbrake cylinderr between eacch brake and thhe HCU.

Liang Chu et al. / Energy Procedia 16 (2012) 1685 – 1692

Author name / Energy Procedia 00 (2011) 000–000

Fig. 2. Schematic diagram of HIL test bench

B. Vehicle dynamic modeling To validate the traction control algorithm been developed, vehicle model is constructed under the environment of MATLAB®/Simulink®. The vehicle coordinate system is a vehicle-fixed right-hand orthogonal axis system. The system includes a seven-degree-of-freedom dynamics model describing the motion of vehicle and wheels. A brief summary of this vehicle model and engine and tire models are presented hereinafter. 1) Engine model The character of engine can be described by the steady state character and the transient state character. The steady state of engine can be modeled by lookup table and the transient state can be modeled by a first order system with time lag, as equation 1 shows:

Me = Ms

1 e −T2 s 1 + T1s

(1) Where: Ms is the steady state torque of engine, Me is the transient torque of engine, T1 and T2 are time constant.

1687

1688

Liang Chu et al. / Energy Procedia 16 (2012) 1685 – 1692

Author name / Energy Procedia 00 (2011) 000–000

Fig. 3. Structure of simulation model of HIL test bench

2) Vehicle model • ⎛• ⎞ M ⋅ ⎜V x −ψ ⋅Vy ⎟ = ∑ Fx ⎝ ⎠

(2)

⎛ ⎞ M ⋅ ⎜V y +ψ ⋅Vx ⎟ = ∑ Fy ⎝ ⎠ •



(3)

••

I z ⋅ψ = ∑ M z

(4)



I i ⋅ ω i = Fxi * rdi − Tbi − T fi

∑ Fx = ∑ Fxi ⋅ cos δ i − Fyi ⋅ sin δ i

)

(6)

∑ F = ∑ (F

)

(7)

4

(

(5)

i =1 4

y

i =1

xi

⋅ sin δ i + Fyi ⋅ cos δ i

• Where: • V y is the lateral acceleration of • mass of vehicle, V x is the longitudinal acceleration of vehicle, M is the vehicle, ψ is the yaw rate of vehicle, Iz is the rotational inertia of vehicle about z axle of vehicle in vehicle reference system, Fxi and Fyi are tire forces in the x direction and y direction in wheel systems.

3) Tire model

⎧C S l 2 + μ x Fz (1 − 3ln2 + 2ln3 ) Fx = ⎨ s s n ⎩μ x Fz

S s < S sc

⎧⎪Ca S a ln2 + μ y Fz (1 − 3ln2 + 2ln3 ) Fy = ⎨ ⎪⎩μ y Fz

S a < S ac

S s ≥ S sc

S a ≥ S ac

(8)

(9)

1689

Liang Chu et al. / Energy Procedia 16 (2012) 1685 – 1692

Author name / Energy Procedia 00 (2011) 000–000

⎧⎡ ⎤ 2 ⎛ 1 2 ⎞ 3 ⎪⎢Ca S a ⎜ − + ln ⎟ + μ y Fz ⎥ ⋅ ln ⋅ l M z = ⎨⎣ ⎝ 2 3 ⎠ 2 ⎦ ⎪0 ⎩

S a < Sc Sa ≥ Sc

(10)

Where:

1 3μFz ln =1 − S n μF S Sc = 3 z Cs Sn =

(Cs S s )2 + (Ca S a )2

μ = μ x2 + μ y2 C S ac = s S sc2 − S s2 Ca 3.

Fuzzy pid controller for traction control

Traction control system mainly controls the output torque of engine and braking torque of driven wheels to prevent the driven wheels from spinning. The target of the system is to control the slip ratio of driven wheels to the region where the coefficient of friction between tire and road is at its optimum. This can improve the acceleration performance and the lateral stability of vehicle driving on slippery road. Engine throttle control

delt_ kh Va_act

Engine Fuzzy PID controller

Vehicle plant

Vat Signal processing

Vref

Te

Traget speed calculation

Vwt

delt_p Brake Fuzzy PID controller

Wheel speed

Tb Brake system

Wheel speed sensor

Fig.4. Structure of traction slip controller

The structure of the fuzzy PID traction control algorithm validated in the HIL simulation is shown in figure (4). As the slip ratio at which the coefficient of friction between tire and road is at its optimum varies with adhesion coefficient of the road and the vehicle speed etc. The first thing in TCS control is to calculation the target speeds. The target speeds including target speeds of driven wheels and target speed of drive shaft. The target wheel speeds can be calculated by vehicle reference speed and the target slip ratio, the target speed of drive shaft can be calculated by the average of the target speeds of driven wheels. As mentioned hereinabove, the target slip ratio varies with the road adhesion coefficient, so it is of advantage to determine the target slip ratio as a function of the two factors. As following equations show:

Vwhl_tar = Vref ⋅ (1 + λtar )

(11)

1690

Liang Chu et al. / Energy Procedia 16 (2012) 1685 – 1692

λtar = f (Vref , μ )

Author naame / Energy Prrocedia 00 (201 11) 000–000

Vds_tar = (Vwhl _ tar _ L + Vwhl _ tar _ R ) 2

(12)

(13) Where: Vwhl_tar is the target t speed off driven wheel; Vref is vehiccle reference speed; s λtar is taarget slip ratio o of driven whheel; Vds_tar is the target speeed of drive shaft; s Vwhl_tar_LL and Vwhl_tar_RR are target sppeed of driven n wheels on thhe left and righht side of vehiicle.

Fig.5. Structuree of adaptive fuzzzy PID controllerr

Fig. 6 Surface view of the param meters of fuzzy PID P controller

After the target speedss been calculatted, a fuzzy PID controller is used to conntrol the enginne throttle and d d wheels. The differennce between target t drive shhaft speed andd actual drivee the braking pressure of driven e throttlee controller, where w the actuual drive shafft speed is thee shaft speed is input to the fuzzy PID engine a driven wheel w speeds. The differencce between tarrget wheel speeed and actuaal wheel speed d average of actual of each drivven wheel is seend to the fuzzzy PID brakinng controller. As shown in figure (4). Thhe structure off fuzzy PID controller c is shhown in figurre 5; the errorr and its deriv vation are sentt to fuzzy reaasoning block,, from which the parameters of PID conntroller are decided. The su urface views of o the parametter KP, KI, KD and e, Δe are shown in figgure 8.

Liang Chu et al. / Energy Procedia 16 (2012) 1685 – 1692

4.

Author name / Energy Procedia 00 (2011) 000–000

Simulation results of fuzzy pid traction control algorithm on test bench

Figure (7~9) show the simulation results of fuzzy PID traction control algorithm on the test bench. The simulation is about the maneuver of straight line acceleration for stop on low μ road. As both driven wheels show the tendencies of spinning, braking control of driven wheels and engine throttle control are activated. 10 speed of front left wheel speed of front right wheel vehicle reference speed

9 8

speed(m/s)

7 6 5 4 3 2 1 0

0

1

2

3

4

5

6

7

Time(s)

Fig. 7. Wheel speed and vehicle reference speed with TCS 2 brake pressure of front left wheel brake pressure of front right wheel

1.8 1.6

Pressure (Mpa)

1.4 1.2 1 0.8 0.6 0.4 0.2 0

0

1

2

3

4

5

6

7

Time(s)

Fig. 8. Braking pressure during TCS control

As shown in figure 5, when driving off, both the left side and the right side driven wheel start to spin, this means that the engine output torque at the driven wheel, i.e. the drive force, is greater than the maximum force the ground can provide, so the engine throttle control decreased the opening amount engine throttle, as shown in figure 7. Besides, as the vehicle reference speed is low, so the braking controller of both driven wheels is triggered, as shown in figure 6. The speeds of driven wheels are controlled to their target values.

1691

1692

Liang Chu et al. / Energy Procedia 16 (2012) 1685 – 1692

Author name / Energy Procedia 00 (2011) 000–000 100

Engine throttle openning degree

Throttle(%)

80

60

40

20

0

0

1

2

3

4

5

6

7

Time(s)

Fig.9. Engine throttle during TCS control

5.

Conclusion

In order to validated the fuzzy PID traction control algorithm proposed, a HIL test bench based on xPC Target™ product of MATLAB® is introduced. The test bench consists of the hardware part and the software part, the hardware part is based on the xPC Target™ product of MATLAB®, the software part of the test bench, mainly the vehicle model, is created in Simulink. The validation of the proposed algorithm is implemented on the test bench and the results show that the spin of driven wheels can be controlled effectively and the acceleration performance can be improved when driving on slippery road. References [1] William Blythe, Terry D. Day: “Single Vehicle Wet Road Loss of Control: Effects of Tire Tread Depth and Placement”, SAE Paper No. 2002-01-0553, March 2002. [2] Soo-Jin Lee and Young-Jun Kim, Kihong Park, Dong-Goo Kim “Development of Hardware-in-the-Loop Simulator and Vehicle Dynamic Model for Testing ABS” SAE Paper No. 2003-01-0858, March 2003. [3] Huiyi Wang: “Hardware-in-the-loop Simulation for Traction Control the Debugs of its Electric Control Unit” SAE Paper No. 2004-01-2056, May 2004. [4] Huiyi Wang, Bo Gao, Jian Song “Approach of debugging control laws of ABS combined with hardware-in-the-loop simulation and road experiment”. SAE Paper No. 2001-01-2729, November 2001. [5] Liang Chu, Wanfeng Sun, Yong Fang, Mingli Shang, Jianhua Guo: “Development of a Test Bench for Integrative Evaluation of the Pneumatic ABS TCS Performance”. Vehicle Power and Propulsion Conference, 2009. VPPC '09. IEEE Press, Sept. 2009,pp.1079-1084, doi:10.1109/VPPC.2009.5289729 [6] Jianlong Zhang, Deling Chen and Chengliang Yin “Adaptive Fuzzy Controller for Hybrid Traction Control System based on Automatic Road Identification”. Proceeding of the 2006 IEEE International Conference on Automation Science and Engineering Shanghai, China, October 7-10, 2006, PP 524-529. Doi: 10.1109/COASE.2006.326936 [7] Jian Pei,LiMing Zhao, DeJun Wang and Liang Chu “Fuzzy PID control of traction system for vehicle ” Proceedings of the Fourth International Conference on Machine Learning and Cybernetics, Guangzhou, 18-21 August 2005, pp. 774-777, Doi: 10.1109/ICMLC.2005.1527048. [8] Quan-Zhong Yan, Faisal Oueslati, Jim Bielenda and John Hirshey: “Hardware in the Loop for a Dynamic Driving System Controller Testing and Validation” .SAE Paper No. 2005-01-1667, April 2005. [9] Andreas Riedel and Alexander Schmidt: “Testing Control Systems of Trucks and Truck-Trailer-Combinations with Hardware in the Loop – Very Real Tests in a Virtual World” .SAE Paper No.2001-01-2768. November 2001.