Optimal Torque Vectoring Control for Autonomous Vehicles Considering Ride Comfort

Direct Yaw Moment Control (DYC) has emerged as a crucial technique in enhancing vehicle stability and handling, particularly given the growing flexibility in electric vehicle powertrain and chassis configurations. The ability to control the yaw moment through torque vectoring offers significant advantages in vehicle dynamics, especially during complex steering maneuvers. However, a critical challenge lies in balancing the influence of steering input and torque vectoring, as this balance is often predetermined by system parameters, which may not adapt optimally to varying driving conditions. In this paper, we propose a novel integrated approach that combines steering control with force distribution on the rear axles of an autonomous vehicle equipped with dual rear motors. To achieve this, a dual-track vehicle dynamic model is developed, which incorporates simplified roll dynamics. This model takes into account the steering angle and individual rear tire forces as primary inputs, allowing for a more nuanced control strategy. The core of our approach is the formulation of an optimization problem that determines the optimal timing and magnitude of Torque Vectoring activation. This ensures that the torque from the rear motors not only enhances the vehicle's steering capability but also contributes to overall stability and safety. Furthermore, we introduce a comfort index designed to evaluate the driving trajectory generated by the upper-level planner. This index serves as a critical metric for assessing ride comfort, which is a key consideration in autonomous vehicle operation. To validate the proposed method, extensive simulations and real-world experiments were conducted. The results demonstrate the feasibility and effectiveness of our approach in improving both vehicle dynamics and passenger comfort, making it a promising solution for future autonomous vehicles.