Vehicle Control Of Four-Wheel Driven Micro Electric Vehicle

With the challenges of energy and environment crisis, battery electric vehicle isconsidered to be one of the solutions in the transportation field due to its advantages inhigh efficiency and zeros emission. Micro electric vehicle is born with the advantage inenergy efficiency thanks to its lightweight body, which leads to a lower requirement forthe battery capacity and thus a less cost increase. Therefore, micro electric vehicle ismuch likely to be the first type of battery electric vehicle accepted by the market. Basedon the development of vehicle control system of micro electric vehicle, this dissertationfocuses on the system configuration optimization and energy management, whichconsists of torque distribution, multi-sensor fusion and target torque planning.Firstly, the in-wheel motor driving is selected to be the system configuration basedon analysis and comparison of different system configurations of battery electricvehicles. Then the parameter designs of in-wheel motor and power battery is studiedusing simulations in Matlab/Simulink to meet the design requirements of the microelectric vehicle in usability, energy efficiency and cost according to China’s actualconditions.A vehicle control system is developed for the micro electric vehicle, including theTTCAN-based distributed vehicle control system design and the hardware and softwaredesigns of the vehicle control unit and the motor control unit. Vehicle tests have verifiedthe functionality and reliability of the vehicle control system.To address the problem that the conclusion in literature in the study of torquedistribution for in-wheelmotor driven electric vehicles disagrees with the experimentresult, a comprehensive efficiency model of in-wheel PMSM motor is developed，based on which the optimal solution for the torque distribution problem is theoreticallyderived using analysis from the power loss minimization aspect. And the conclusionproposed by this dissertationis verified by vehicle tests.In order to provide information of vehicle, road and traffic conditions to the targettorque planning, a multi-sensor system which consists of GPS, IMU, electric drivingsystem, radar, and traffic information system is developed. And the electric drivingsystem is integrated in the multi-sensor system for the first timeto make full use of the advantages of in-wheel motor driving. The multi-sensor fusion algorithm is studied,which could provide estimations of the position, velocity, acceleration, attitude andangular velocity of the vehicle, road gradient and vehicle mass. Test results show that,the system positioning accuracyof2m and10m could be achieved with and withoutGPS respectively, and the vehicle mass estimation accuracy could reach2%.Based on the information from the multi-sensor fusion system, target torqueplanning is studied to optimize the profiles of target torque and velocity when thevehicle is driving in urban roads with variable road gradient and predictable traffic lighttiming. Model predictive control and dynamic programming are employed to derive alocal optimal solution to the problem of target torque planning. By transforming thetwo-dimensional dynamic programming into one-dimensional dynamic programmingusing Lagrange multiplier method, the computation time could be reduced by90%.Simulation and vehicle test results show that, the energy efficiency of the micro electricvehicle could be improved by9~10%using target torque planning.