ACTIVE FRONT STEERING DURING BRAKING PROCESS

CHINESE JOURNAL OF MECHANICAL ENGINEERING●64.Vol. 21, No. 4, 2008DOI: 10.3901/CJME 2008 04.064, available online at www. cjmenet.com; www.cjmenet.com.cnACTIVE FRONT STEERING DURINGCHEN DelingBRAKING PROCESSCHEN LiYIN ChengliangAbstract: An active front stering (AFS) intervention control during braking for vehicle stability ispresented. Based on the investigation of AFS mechanism, a simplified model of steering system isestablished and integrated with vehicle model. Then the AFS control on vehicle handling dynamicsZHANG Yongduring braking is designed. Due to the difficulties associated with tbe sideslip angle measurement ofSchool of Mechanical Engineering,vehicle, a state observer is designed to provide real time estimation. Thereafter, the controller with theShanghai Jiaotong Universityfeedback of both sideslip and yaw angle is implemented. To evaluate the system control, the proposedShanghai 200240,ChinaAFS controlled vehicle has been tested in the Hardwarein-the-loop-simulation (HILS) system andcompared with that of conventional vehicle. Results show that AFS can improve vehicle lateralstability ffectively without reducing the braking performance.Key words: Active front string(AFS) Handling stability Yaw rate Sideslip angleHardware-in-tbe-loop-simulation (HILS)using both the yaw rate and sideslip angle is designed in section 2.0 INTRODUCTIONMoreover,a state observer is aplied tostimate the sideslingle.Hardware-in-the-lop-imulation (HILS) test based on theHandling stability is one of the most important aspects ofnonlinear model for AFS intervention control during braking arevehicles. In the process of asymmetric braking, additional yawprovided in section 3. Finally, conclusions on the AFS controlduring braking are drawn in section 4.momentvll beproouccoarand result in vehicle instability. Mostpreset sabilly cnharcerment tchniquesp such as dreet yaw 1 SYSTEM DYNAMICS MODELING .control (DYC) and electronic stability program (ESP) use theavailable braking or traction components in aiding driverThe output of the seeing system is the steering angle, whichDYC utilize tyres as longitudinal forces actuators to generate yaw is the input of the vehicle dynamics model. Thus, the systemmoment by the longitudinal force-differential between the left and dynamics model includes both the AFS and the vehicle models.right wheels [B5]. ESP maintains vehicle stability through1.1 AFS modelreleasing braking and multiple cycles of the wheels, which is aThe AFS system is composed of four basic elements: atradeoff between stability and braking distance !Both the technologies utilize longitudinal force to maintainStenngig column, a steering rack, a permanent magnet synchronousvehicle sabili However, it results in decresing longiudinal motor (PMSM) and a planetay gear se, asilusated in Fig. 1.fore, which may be undesired, such as in emergeney braking The PMSMs cortroled to provide the desired ssisant anglewhere maximal braking forces are needed. Active front sering aoring to the diver's demand and veloeity. The asistastt angle(AFS) system can improve vehicle sabili by rgulaing the front and the driver's operation are superimped to the road seerinstering angle. Then aditional lateral forces can be produced and wheels through a planetary gear set.applied to the front wheels, independently of the driver, toappneaetocompensate the unbalance moment7. In this way, AFS alsoSteeingimposes the direct yaw control on the vehicle without decreasingVelocitybraking forces. As a result, more braking forces can be appliedAnele se_cand braking distance will be shortened. On the other hand, the tyredynamics of most normal driving situations are within linearregion, where AFS systems can work more efectively than DYCPMSM- ( Torque controland ESP (8). Although most present AFS researches are dedicatedSteering column'to improving vehicle steering response, tbe value of AFS oDstability enhancement cannot be neglected.FrontwheelTo improve vehicle stability,y, most researches used only theTe rodPinon and rackyaw rate as feedback due to difficulties associated with sideslipangle measurement [9 10. Theoretically, the vehicle stability ofFig. 1 Structure of the active stering systemlateral motion is described by yaw rate and sideslip angle. TheIn this model, the steering wheel angle is taken as the input.yaw rate alone cannot determine the vehicle dynamic charactersTo simplify the model, all steering mechanisms are supposed nigidcompletely. Additionally. drivers are more sensitive to the vehicleexcept the steering column and steering linkages. The torsionalsideslip than yaw rate [0. Therefore, it is better to utilize both ofstiffness of the steering column and the equivalent torsionalthem to improve the vehicle stability. Many attempts have beenstiffness of steering linkage are combined and denoted as kp 中made to estimate the sidesip angle. For example, Ref. [11]proposed an estimation method based on lateral acceleration. InThen, the AFS system dynamics can be described asRef. [12], the combination of global positioning system (GPS) andJinertial navigation system (INS) sensors was used in sideslip. +-=|8.+(B。+B)6。estimating. However, it is expensive to implement.中国煤化工、In the present paper, the AFS intervention during braking(1]process for the stability improvement is proposed. The paper isYHCNMHG)=Torganized as follows. Firstly, the AFS system model integratedwith vehicle dynamics is described. Then the feedback contollerJ8,+B.S,+k.C,G8-8-c =M,. (2)Reccived February 26, 2007; received in revised form May 15, 2008;NGaceped July 10, 2008CHINESE JOURNAL OF MECHANICAL ENGINEERINGJw,=-RF+T i=fl, f,r,π(6)included) moment of inertiawhere Fm,Fm,Fr,Fr(F)- -Longitudinal forces on the frontB-Steering column and rack dampingleft, front night, rear lef, rearGear ratio of motorright tyresG,一 Gear ratio of steering mechanism systemF.FFN,FEm-Lateral forces on the front left,k。一-Torsional siffness of steering systemfront right, rear lef, rear rightN - - -Number of PMSM pole pairstyresT一Difference between the driving-Motor column moment of inertiatorque and the braking torqueB.- -PMSM dampingapplied to each wheel- Steering angle produced by the PMSMm一Vehicle massJ.一-Wheel moment of inertia about seering axleJ一Vehicle moment of inertiaFront wheels dampingw -Track width8一Steering angle of the front wheelJ的一Moment of inertia of tyre about8一-Steering wheel angle provided by driverits rotation axisa.一-Wheel rotational speedT。一Torque provided by the PMSMR一Efctive tyre radiusM,'一Tyre self-aligning moment, in the simplified linearγ - Yaw rate, the yaw velocity ofmodelthe chassisM, =2kd(β+二r-8)a,b一- Distance from CG to front andrear axleEq. (1) describes the PMSM dynamics. The left sideV,VyLongitudinal and lateral velocity.represents the effect of the inertia, damping and distortion of theThe longitudinal and lateral forces acting on the wheels arePMSM. The right side is the torque applied by the PMSM. In the depended on dynamic performance of the tyres. The Dugoffsimplified system model, the torque is assumed the direct output model, which requires the smallest computational effort, isof the controller. Front wheel dynamics is described by Eq. (2),introduced to calculate the forces generated by thewhere sideslip angle B, yaw rate r and 8 r are included.According to the nonlinear vehicle model, each wheel has anindependent slip angle1.2 Vehicle modely +ayIn this research, a nonlinear model neglecting roll and pitchan(an)= 8 - arctan口FWr12(7a)motions is used. The nonlinear model accommodates steering andindividual wheel braking actions (Fig. 2).a(an)= arctanby-v, )(7b)(y,FWr12)Moreover, the longitudinal wheel slip is defined as兴且[。一RaRa, 1m(的, +v,r)=(Fn+ F)osδ +(Fw +Fm)+k一Cornering stifness of each tyre(Fn+ F)sin8r(4)中国煤化工1tyreYaw motionCNMHG. also expressed asJj =[(Fm +Fm )cos8&s +(Fn +Fn)sin&;]a-(Fm +F)b+[(-Fm+Fm)cos&-Fu+F_]W/2(5)E,--Road adhesion reduction factorFriction cofficient of the roadWheel rotational motion●66.CHEN Deling, et al: Active front seeing during braking processReference modelThe instantaneous vertical load acting on each wheel is thesum of the static tyre load, load transfer due to longitudinal andlateral acceleration. The load of each wheel6] can be expressed asFmn(Fa)=2a+6 mgb-mi.h杜w b2F.1(11a)ContollerAFSVehicle t十.βFig.3 AFS system cotrolF(F)=zmga + my,h土(11b)2(a+b)According to the reference model, the target yaw ratey= /(a+b)δwhereF, = mv,(β+r), F, is the seering comering force, βisYo= 1+40)广the sideslip angle of vehicle. Fan,Fa,FuvFm are the verticalwhere A is the stability factor indicating the charateristics ofloads of front left, front right, rear lef, rear right respectively; hvehicle steering; 8 is the steering angle input by the driver,is the height of gravity center.ris the variable steering ratio. In the braking condition with2 FEEDBACK CONTROLLER DESIGNstraight lane, there should be zero yaw rate and no lateraldeviation occurs. Then the desired yaw rate and sidesip are bothIn the process of controller design, tradeoffs are usually zero.necessary between modeling accuracy and mathematicalTo implement the feedback control scheme described above,complexity. An extremely accurate model often requires large accurate information on sideslip angle and yaw rate are required.computational effort that might be too excessive for system The yaw rate signal can be measured with a gyroscope whileanalysis and controller design purposes7. A high-order nonlinear:desipmeasurement requires expensive speed over groundmodel can be linearized to obtain a simplfed o. According tosensors. Aearly researches [. 18 19, a simplified vehicle model is generally estimate the sideslip angle by the measurable information.derived by linearizing the nonlinear model with respect to theThe standard observer structure is shown in Fig. 4, the stateoperating point of interest. In the model, both the front and rear space equations are presented asl01wheels are lumped at the vehicle center line. The dynamics ofsteering system and wheel rotation, the suspension operation and[()= A()+ Bu()- H[j()-x(0](14)the tire lateral deviation created by the road tangential force are[j(1)= C&()neglected. Assuming smal perurbations, a linear rleaionshiplinear relationshipbetween lateral force and slip angle can be established. Then the Then the estimated vector is given bytwo degrees of freedom (2-DOF) linear model, which describesvehicle lateral dynamics well in the linear region, is employed to()=(A-. HC)<(1)+ Bu()+ Hy(I)(15)facilitate the controller-design process. Taking sideslipangle β and yaw rate r as the states of the system, the 2-DOFmodel can be described as2(k+k) 2(ak, -bk) _1|(-244tmy,mv;δ (12)2(ak-bk) 2(a2k +b'k5)HtJ心(J,Fig. 4 Structure of the state observerwhere k.kq are the cormering sifness of front and rear wheels.The vehicle representation for estimation is identical to theWhile a vehicle is runing, actual and target vehicle simplified 2DOF model introduced previously. But the originaldynamics may be dfferent when disturbance is aplied to the state vector x of system dynamics model should be rewritten astyres or the road condition is changed suddenly. To improve thex and x,where x is the estimated vector which can not bevehicle handling sability, the yaw rate and the sidesip anglemeasured directly, 上is the vector can be measured withshould be controlled to follow their target values.As in Fig. 3, the reference model provides the target valuessensors. The new equations of state space can be expressed asof vehicle state variables according to driver input and vehiclevelocity. The 2-DOF model is used as the reference model. rs,Ps()(x 1)(&)rget values of yaw rate and sidesip angle. Both of themare employed to regulate the motor compensating angle for thevehicle stability. Then the feedback control law for the active wherex,=(8。&, B)",x=(8。8, r川， u=(8。 T.)",steering vehicle can be gottenand y is the measured vector available from vehicle sensors.T=K(r -r)+K(β- B)+ K.8。(13)R。R中国煤化工where K,Kp一 Feedback cofficientsof r and β,respectivelyMYHCNMHG. 2k.dK。一-Feedfoward coeficient, which can beJ。gotten according to the variable steering2(k+k)ratio and the PMSM parametersmv,CHINESE JOURNAL OF MECHANICAL ENGINEERING●67●-大kG.Jm(NG,)尸+J NGJm +J (NG,)k,G2kd+k.C2kdaAb=NG,JJv,.2(ak, -bk)0mv,mv2l001o。2(ak-bk)。-2ak 2(a2k.+b352)(a) Laboratory setupv.JSteeringSideslip. yaw rate,velocityNGJL+J(NG,) J_+L(NG,)P, AngleECU0 0)signalkGmotorPC andB,=B,=|0 0J。(0 0)DA cardsMP- 2Steering links(000100)Displacement sensor-a的B=(B) c000010|(b) Scbematic diagram of experimental test component000001)rig.5 Hardware-io.-bhe loop simulation systemThe state equation of the estimated part isThe parameters characterizing the vehicle model are listed inx=Ax.+Auy+ B,u .(17)Table. The simulation involves two steps. Firstly, to make sure thereliability of the HILS results, the vehicle model is venified. ThenThe state equation of the measured part iscomparisons are performed to clarify the efctiveness of the AFSjy=Ax +Awxp+ B.u .(18)intervention control.Table Vehicle parametersParameterValuebe measured with sensors. Then the state observer can beexpressed asVehicle mass m/kg1 080Distance from CG to frotaxle a /m1.138文=(4- HAw)元+ Amy+ B.u+ Hfy-Distance from CG to rearaxle b /m1.321H(Any+ B.u)(19) Veticle mass momet of ieria J, /0kgrm)1426To avoid using j directly, an altemnative vector is usedPneumatic trail of front whees d /血Torsional siffness of sering system ks /(N*mrad)65.2[位=(A.-HA.)z. +(B,- HB,)u+[(4m - Ha.)H+Ab- HAs]y(20)PMSM load momeat of inertia JL /(kg-m)0.04Steering column and rack damping B /Nomsrnd )0.022 s(x=z +Hy0.000 sThenδm,o,β can be estimated with the measured vectorsPMSM damping Bm /(Nmsrad~)625by the state observer. Both the yaw rate and sideslip angle signalsWheel mass moment of inertia about axle Jw /(kgm)0.568can be sent back as the feedback of the AFS control.Front wheels damping B. /(N-msrad~)0.013 SIMULATION AND EXPERIMENTMotor gear ratio G%A hardwre-in-loop simulation (HILS) system is built as inSteering system gear ratio G14.4Fig. 5 to evaluate the performance of the proposed AFS system.Track width W /m432The HILS system is composed of hardware (steering systemHeightofCG h/m0.49mechanics, ECU, motor, data acquisition card, sensors, PCs) andsoftware (vehicle simulation model as Eqs. (3)-(11) in MATLAB).Longitndinal stifes of one tyre k, /(N/unit sli)53 018/54 526The ECU receives the driver steering wheel input from theComering sfines of one tyre k /(Nrad~")-34 500/ -35 500steering angle sensor, the yaw rate and sideslip angle informationfrom the simulation of the vehicle model in PC. And the motorRoad adhesion reduction factor 6, /(s-m' )0.015generates the torque according the command from contoller andEfeci中国煤化工0.265its angle is measured by the potentiometer. Then the steeringTyre |1.070.85linkages are moved with the displacement measured by the sensor.:YHCNMHGlkem)With the sensor signal, the vehicle model resides in the PC iscomputed, then yaw rate and sideslip angle signals are sent back3.1 Model verificationto the ECU as the feedback.To venify the models, AFS system without motor assistance●68.CHEN Deling, et al: Active front seering during braking processare simulated and compared with experiment. The experiment is80~一Front left wheelcarried out on the conventional vehicle without any stability-- - Rear left wheelimprovement technologies.Both the simulation and experiment营60-.... Front right whee!are carried out in the same conditions.-- - Rear nght wheelThe experimental input (braking pressure, which gives theindication of brake torque as: T= KP(1)，whereK is brakingcficicnt and P() is braking pesure吵is used as input to themodel simulation, as in Fig. 6a. The wheel slips are contolled to层200.2 during this braking condition. It can be see that the pressuresof each braking cylinder are not uniformity. Therefore, additionalyaw moment will be produced and result in vchicle instability.As the sideslip angle and the lateral acceleration have theTime dtsrelation as:a, =β+v;(β+y). The lateral aceleration can(a) Braking pressure70prepresent the sideslip dynamics and is measurable in thexperimental vehicle. Therefore, the lateral acceleration is usedinsted of the sdsis angle in the rsuls comrsrsrirori,pThe vehicle is driven at an initial velocity of 60road friction coefficient of 0.3. Figs. 6b-6d are the brakingdistance, yaw rate and lateral acceleration responses, respectively.The simulated vehicle braking distance is shown to be a veryclose match to the experiment. The simulation results of yaw rate.... Conventionaland lateral acceleration exhibit similar tendency and peak values- - - Expenmentalwith the experiment although some bias exist. The bias betweenthe simulation and experiment is mainly due to the test methodand the exterior disturbance. The acceleration measurements areTime teasily affected by disturbances such as road bank angle,(b) Braking distancesuspension roll. Therefore, biases exist inevitably. In aionall,0.05to verify the effectiveness of AFS, it is assumed that the driver..Conventionaldoes not provide any corrections to the vehicle instability but0- - Expenmental。keeps the seering wheel straight. In the simulation, the sterin日-0.05wheel angle is set to be zero to meet the assumption.Wilennhkeep the steering wheel sill Generally, the effects are the same,-0.15-wbut not always exactly consistent. For example, when theacceleration drops, the originally force is used to keep the steeringwheel still for the delay of the driver and the steering mechanics.-0.25Thus, the force used to balance the braking disturbance is larger(absolute value) than expected and makes the lateral acceleration-0.30drop slowly. Both the test method and the disturbances result inTime /sthe error between the experiment and simulation during the later(C) Yaw rate of the vehicleperiod of the test. But the peak and trends represent the main0.5厂performances of system. Moreover, the bias will weaken when the..o.ventionasimulations are taken in the same test method.. ExperimentalConsidering both the yaw rate and lateral accelerationcomparon, the models are effective and can be used to evaluatethe performance of the proposed active steering system.3.2 AFS control experimentThe braking maneuvers are performed under two dfferentworking conditions: braking on a low-friction road and emergencybraking on split adhesion road. Comparisons are performedbetween the AFS controlled vehicle and conventional vehicle. In-3.0-士order to verify the stability enhancement by the AFS interventionTime "s .control and reduce the bias, the steering wheel is fixed by themechanics in the HILS system.(d) Lateral acleraion3.2.1 Braking on lowfiction roadFig.6 Model verificationFirstly, the vehicle is tested under the same condition as the3.2.2 Emergency braking on split adhesion roadmodel verification. The braking pressure is used as the inpu. Fig.It is expected that the vehicle can stop as quickly as possible7a is the braking distance. The braking distance of the AFSwithout locking the wheels in emergency braking. The maximalcontrolled vehicle is a lttle longer than that of the conventionalvenicle, wWhch shows the braking perfomane hs not been braking fres are aplied to each whee wih dferet ahesionbeavily reduced. Figs. 7b-7c are the yaw rate and the lateralconditions on the left and right sides. Thus, a yaw moment isacceleration, respectively. The vehicle with AFS interventiongenecratre中国煤化工ie with higher fictioncontrol is scen to achieve lower peak lteral aeleation and yaw cofficieFe to coret the lanerate in response to the asymetric braking. Thus, conclusion can change.YHCN M H G to keep the brakingbe drawn that the sability is improved duing braking with the forces as maximal as pssible to stop the vehicle. Therefore, steerAFS intervention and Fig. 7d is the compensating torque providedshould be used. The AFS can correct the vehicle instabilityby the PMSM of AFS.without driver's compensation.CHINESE JOURNAL OF MECHANICAL ENGINEERING●69●35p60_30鼻4020a 10...Conv.ntiona.t,.... Conventional- AFS interventionAFS intervention6Time D!:Time tme ts(a) Braking distance(@) Braking distance0.10pConventional0.055--0.05-Mwwm1.5--0.102.0F言-0.15st-- Conventional-0.20--0.25 2一↓Time t!sTime Uts(b) Lateral acleration(b) Yaw rate5厂05r. Conventiona!日-05王-0.05-1.0F-0.10-邑虽-0.15-2.0....on.ventional之54Time i'sTime ts(C) Yawrate(Q) Lateral acceleration2.00.05厂. Conventional- AFsintevenier0.04-5- |美0.奚0.012-.56一 g-0.05之4Time th(d) Compensating torque provided by PMSM(d) Sideslip angleFig. 7 Simulartion of brnking on lowfiction roadThe emergency braking maneuvers is tested with the initialvelocity of 60 km/h. The road friction coefficient of left side is 0.3,and right side is 0.6. Fig. 8a is the braking distance, which showsthat the braking performance has not been heavily reduced. Figs.8b-8d are the lateral acceleration, yaw rate and sideslip angle,respectively. Compared with the conventional vehicle, the yawrate and latera! acceleration of the AFS controlled vehicle islargely reduced. The yawrate，sideslip angle and lateral中国煤化工edgtuetacceleration responses of the proposed controlled system have lessMHCNMHG寸二 ↓overshoot and achieve less stabilization time. Thus, the vehiclewith AFS can recover the stability quicker than conventional one.(e) Compensating torque provided by PMSMThis is due to the compensating torque provided by the PMSM, asshown in Fig. 8e. .Fig. 8 Simulation of emergency braking on split adhesion road●70●CHEN Deling, et a山Active front seering during braking processAs demonstrated by the HILS resuts, the AFS can improvesystems Tchnology, 2005, 13(6); 965 -975.the vehicle stability and does not weaken vehicle longitudinal[13] YU Fan, LIN Yi. Vchicle system dynamics [M]. Beijing: China Machineperformances.In addition, when ESP is used in this emergencyPress, 2005. (in Chinese)braking on split adhesion road, the braking force of the right side [14]GUVENC B A, ACARMAN T, GUVENC L. Coordination of seeringwheels will be reduced to keep the vehicle stability. In this way,and individual wheel braking achuated vehicle yaw stability control[CWthe braking distance will be increased. According to the relativeIEEE Industrial Electronics Society. IEEE Proceedings of Inelligentresearch, the braking distance of ESP control will be 47.185 mVehicles Symposium, June 9-11, 2003, Columbus, Ohio. USA: IEE,through simulatin, while the braking distance of our AFS2003: 288 -293.intervention control is 29.981 m.[I5] DUGOFF H, FANCHER C, SEGEL L. An analysis of tire tractionproperties and their influence on vehicle dynamics performance[CW4 CONCLUSIONSSociety of Automotive Engineers. Procedings of FISITA IntecnationalAutomobile Safety Conference, May 1970, Detroit, Michigan. USA:(1) The stability improvement by active front steering (AFS)SAE, 1970, Paper No. 700377: I 219-1 243.intervention during asymmetric braking has been discussed.(2) The AFS feedback contoller is designed using both yaw[16] ALEJANDRO D, DOMINGUEZ-GARCIA J G.. KASSAKIANJE,etrate and sideslip angle. And a state observer is established toal A backup system for automotive ster-by-wire, actuated by selctiveestimate the sideslip angle to fulfill the contol. The controller isbraking(C]/ IEEE Indusrial Elecronics Society. 35th Annual IEEEtested in the HILS system.Power Electronics Specialits Conference, June 20-25, 2004, Aachen,(3) Comparisons are made between the AFS controlledvehicle and the conventional one on low-fiction road and split (17] ARMSTRONG E. Robust cntolloel dsign for fexible stuctures usingadhesion road. From the comparison, it can be seen that the yawnormalized coprime factor plant deriptins(R] NASA Technical Paperrate and lateral acceleration of the AFS controlled vehicle is3325, 1993.largely reduced by the motor automatically without driver's[18] ONO E, HOSOE S, TUAN H D, et al. Robust stabilization of vehiclecompensation. Thus, the vehicle stability has been improved bydynamics by active front wheel stering contro[(C] IEEE Industrialthe AFS without reducing the braking performances.Elecronics Society. Proceedings of the 35th Conference on Decisionand Control, December 11-13, 1996, Kobe, Japan. USA: IEEE, 1996:References1777-1 782. .[1MAMMAR s, KOENIG D. Vehicle handling impro[19] MAMMAR s, BAGHDASSARIA V B. Two degree ofedomsteeringUJ). Vehicle System Dynamics, 2002. 38(3): 211-242.formulation of vehicle handling improvement by active stering[CW[2ESMAIL ZADEH E, GOODARZI A, VOSSOUGHI G R. Optimal yawIEEE Industrial Electronics Society. Proceedings of the Americanmoment control law for improved vehicle handling[]. Mechatronics,Contol Conference, June 28 30, 2000, Chicago llinols.s USA: IEEE,2003(13): 659- 675.2000: 105-109.[3] FUJIMOTO H, TSUMASAKA A, NOGUCHI T. Diret yaw -moment[20] LIU Bao. Modem control thory [M]. Beijing: China Machine Press,control of electric vehicle based on comering stifness estimation[C]IEEE Industrial Electronics Society. IECON 2005, 31st Annual2003. (in Chinese)Conference of IEEE, November 6-10, 2005, Norh Carolina. USA: lEEE,(21] ZHAO Shuang, SUN Renyun, NING Fankun. Analysis of automobile2005:2626-2 631.ABS algorihm[J]. Machinery Design & Manufacture, 2005(9); 64 65.[4] RAKSINCHAROENSAK R, NAGAI M. Vehicle motion control isues(in Chinese)using micro etric vehicle "Nove!"ICW The 22nd International Bttery,Hybrid and Fuel Cell Electric Vehicle Symposium & Exposition.October 23-28, 2006, Yokohama, Japan. EVS-22, 2006: 1 635-1 646.Biogrephical notes[5] SAVKOOR A R, CHOU C T. Application of aerodynamic actuators toimprove vehicle handling[J]. Vehicle System Dymarnics, 199(32);CHEN Deling is curently a PhD. candidate inSchool of MechanicalEngineering. Shanghai! Jiaouong Uriversty.Chin. HeTresearchinterests[6] AVAK B. Modeling and control of a superimposed steering system[D].Tel: +86-21-34206806; E-mail: chendeling@sjtu edu.cnAlant: Georia Istute of Tchnolog，204.[7] HAC A. DOMAN D, OPPENHEIMER M. Unified contol of brake- andCHEN Li is curendy n associate processor in School of Mechanicalsteer-by-wire systems using optimal control llocation methods(CV1Engineering, Shanghai Jiaotong University, China. She received her PhDSocity of Automotive Enginers. Brake Technology 2006, Api3-6degree from Shanghai Jiaotong University, China, in 2000. Her research[8HE Junie, CROLLA D A,LEVESLEY MC, etal. Coordination ointerests include vehicle dynamics and electronic control, automotiveactive steering, drveline, and braking for integrated vehicle dynamicselectronics.Contolp]. Proc. IMechE, PartD: J. Auonobile Enincring 206220); Tel: +86-21-34206805; E-mail: l.hen@sju.edu.cn1 401-1 421.[9] HOSAKA M, MURAKAMI T. Yaw rate control of clectic vehicle using YIN Chenglang is urenty a pocessess in School of Mechanical Eninerng.steer-by-wire system[C]// IEEE Industrial Electronics Society. 8th IEEEShanghai Jiaotong University, China. He received his PhD degree from JilinIntermational Workshop on Advanced Motion Control, 2004. AMCO4. University, China, in 200 He is n advanced technical consultant of DongfengMotor Corporation (Group) and Shanghai Automotive Industry CorporationMarch 25-28, 2004, Kawasaki, Japan. USA: IEEE, 2004: 31-34.[0] YIHP Serbywire iplicaions for vehidle hadting and sity(以I (Group), Chinag His rearc ites include hading and sabiliy ofCalifomia: Stanford University, 2005.automobile, automotive electronics and Hybrid electric vehicles.[1] HIRAOKA T, KUMAMOTO H, NISHIHARA O. Sideslip angleTel: +86-21-34206323; E-mail: clyin1965@sjtu.cdu.cnestimation and active front steering system based on lateral accelerationdata at centers of percussion with respect to front/rear wels([]. JSAEZHANG Yongis curenty apostdoctorin School Ot Mechaneal Engrneern,include vchicle[12] YIH P, RYU , GERDES J C, Modification of vehicle handingTel: +86-21-34206806; E-mail: blue, _hawk@263.netcharacteristics via steer by-wire[J]. IEEE Transactions on control中国煤化工MYHCNMHG