Intelligent vehicles path tracking control strategy based on sliding mode variable weight MPC

Abstract

Traditional model predictive control (MPC) algorithms often suffer from slow trajectory convergence and non-smooth control increments when applied to complex driving scenarios. To address these challenges, this paper presents a novel trajectory tracking control method for intelligent vehicles based on sliding-mode variable-weight MPC. An adaptive variable-weight strategy is introduced within the MPC framework by integrating a sliding mode control (SMC) mechanism. This approach allows for real-time adjustment of the weight matrix in the MPC objective function based on the lateral displacement error and yaw angle error. This improves the controller's ability to adapt, enhances tracking precision, and ensures stability across different driving conditions. A vehicle dynamics model is constructed with front wheel steering angle as the control input, and the influence of different weighting coefficients on control performance is systematically analyzed. The proposed control strategy is implemented in Simulink and validated through co-simulation with a high-fidelity CarSim vehicle model under a double lane-change scenario. Simulation results show that, compared to conventional MPC, the proposed method reduces peak lateral displacement error by up to 71% and achieves notable improvements in yaw angle and lateral deviation of the vehicle’s center of mass. These results demonstrate the effectiveness of the proposed approach in improving trajectory tracking performance, vehicle stability, and dynamic responsiveness.