Abstract: To address the challenges of low efficiency and significant errors in parameter tuning of prediction horizon for traditional model predictive controllers (MPC), this paper proposes a dynamic horizon adaptive MPC controller integrated with a particle swarm optimization (PSO) algorithm (PSO-MPC). The method establishes a PSO-MPC collaborative framework that dynamically searches for optimal prediction horizon parameters based on the vehicle’s real-time state within each control cycle, thereby constructing a time-varying prediction model to enhance trajectory tracking adaptability. The simulation results show that in the double lane change and lane change scenarios, compared with the MPC method with a fixed prediction time domain, the trajectory tracking error of the proposed controller is reduced by 35% and 37% respectively, verifying that the dynamic prediction time domain optimization mechanism can significantly improve the tracking accuracy and dynamic adaptability of complex trajectories, providing a new technical path for the design of autonomous driving control strategies.
李世豪, 甘树坤, 吕雪飞. 基于PSO参数自适应MPC路径跟踪控制研究[J]. 吉林化工学院学报, 2025, 42(5): 89-94.
LI Shihao, GAN Shukun, LV Xuefei. PSO-Based Parameter Adaptive MPC for Trajectory Tracking Control. Journal of Jilin Institute of Chemical Technology, 2025, 42(5): 89-94.
