Research On Electric-wheel Vehicle Handling And Stability Control Based On Torque Distribution

Electric-wheel vehicle is driven by four in-wheel motors which can save the space ofchassis, increase the flexibility of the vehicle, and provide more operating mechanism forelectronic control units. Because of eliminating the limitation of the traditional mechanicalstructure, the vehicle’s stability rely more on the control, so the vehicle handling stabilitycontrol system. is put forward higher requirements toThe vehicle is more than one degree of freedom nonlinear system, so a hierarchicalhandling stability control structure is designed according to the characteristic of actuatorsredundancy. According to the characteristics of yaw rate and sideslip angle, fuzzy control andPID control algorithm are used to correct the deviation of yaw rate and sideslip angle respectivein a dynamic condition for upper controller, then pass the yaw moment which the controlleroutputs and compensation of steering angle to the lower controller. The simulation results showthat the control strategy of upper controller provides satisfactory performance with vehiclenonlinear under different working conditions, and improves the vehicle handling and stability.The lower controller turn the yaw moment and the compensation of steering angle whichthe upper controller outputs to the four-wheel drive torque by torque distribution technology. Inorder to achieve the optimal torque distribution, the objective function is proposed whichcontains the yawing moment of the tracking error, energy consumption, motor torque pulsation,and the Genetic Algorithm is used to optimize the distribution result online. The simulations ofvehicle limit condition show that the optimal torque distribution method combined with theupper controller, which can improve the handling and stability of the vehicle effectively.Obtaining the accurate vehicle posture parameters in real time is the premise of thesatisfactory handling and stability control performance. To solve the problem of the lowprecision of vehicle posture parameters which are difficult to obtain, the vehicle state observeris designed based on the Extended Kalman Filter algorithm, which observed the yaw rate,sideslip angle, longitudinal velocity online, the simulation results show that the data ofobservation is in the range of allowable error compared with theoretical calculated value, andobservational data is effective.