Maximum friction estimation and longitudinal control for a full in-wheel electric motor vehicle
Résumé
The present paper describes a longitudinal control strategy in time-varying road surface adhesion coefficient conditions, for a vehicle equipped with 4 in-wheel electric motors. The controller task is to ensure two main functions : braking (achieving an ABS-like function, i.e., no wheel skid), and traction control (TCS-like function, i.e., low wheel slip) using the electric motor torques as the unique actuators signal sources. The torque of electric motors is available almost instantaneously, unlike the torque generated by classic internal combustion engines. In addition, information about its values can be measured, starting from the current that passes through the motor, which means that advanced control techniques can be applied directly on the transmitted torque. In order to generate the appropriate anti-skid or anti-slip torque for each wheel, firstly an estimation of the instantaneous friction between the wheel and the road has to be made. Next, the maximum friction coefficient between the wheel and the road is computed with the help of an algorithm based on Dugoff tire model, without prior knowledge of the road conditions. Ultimately, the control strategy will compute a maximum torque that is allowed to be transmitted at the wheels in order to achieve TCS and ABS functions, along with a given reference speed to track. The only measures used in this strategy are the wheel acceleration, the instantaneous torque generated by the electric motors and the chassis acceleration.