Adaptive coordinated control strategy for autonomous vehicles based on four-wheel steering

Abstract

To address the challenges of coordinating path tracking, maneuverability, and stability control in real-time during actual driving of autonomous vehicles, this paper proposes an adaptive coordinated control strategy for four-wheel steering and four-wheel drive autonomous vehicles based on the model predictive control algorithm. First, an approach utilizing a recursive least square method with an adaptive forgetting factor is proposed to estimate and update the cornering stiffness of the prediction model in real-time, aiming to enhance the control accuracy of the controller. Second, the phase plane method is employed to divide the vehicle stability domain, incorporating indicators based on maneuverability and lateral stability to facilitate effective vehicle stability control. Next, the control priority is designed, and the weight of the objective function is adaptively adjusted to achieve the overall coordination between path tracking and handling stability control. Finally, co-simulation of Carsim and MATLAB/Simulink demonstrates that under high-speed and large curvature conditions, the proposed control strategy can further enhance the overall performance of the vehicle. On high-adhesion roads, compared to the general fixed weight strategies, the proposed control strategy has improved path tracking accuracy by 50.57% and 18.41%, respectively. Additionally, on low-adhesion roads, the proposed control strategy still maintains precise trajectory control, with path tracking accuracy improvements of 47.04% and 5.8%, while further enhancing the vehicle’s maneuverability and stability.