Optimal torque vectoring control for distributed drive electric vehicle

Abstract

A novel optimal torque vectoring control (TVC) strategy is proposed in this paper to enhance the lateral stability of a dual-motor rear-wheel drive electric vehicle. The structure of the optimal TVC consists of three parts, i.e., pre-processor, model following controller and post-processor. Unlike the commonly used linear single track vehicle model, an accurate nonlinear vehicle model is built in the pre-processor based on Magic Formula tyre model. The model following controller is responsible for producing the corrective yaw moment by a two-dimensional gain scheduling method related to the vehicle longitudinal velocity and lateral acceleration. This optimal yaw moment controller consisting of the steady-state control law and the optimal feedback control law is developed to compensate the nonlinear property induced by time-varying tyre cornering stiffness. In the post processor, torque vectoring allocation strategies are presented considering the constraints of motor peak torque and tyre friction. Co-simulation results of the CarSim and LabVIEW under two driving manoeuvres (step steering and skid pad track) illustrate that the lateral and longitudinal performance of the vehicle is greatly improved and experimental results of hardware-in-the-loop (HIL) proves that the control system can be well used in real-time.