Dynamic Vector-Velocity Potential Field Model Predictive Control for High-Precision Vehicle Tracking

In the vehicle path tracking domain, traditional model predictive control methods often struggle to maintain high performance in complex and dynamic driving environments. To address these challenges, this paper presents a novel Dynamic Ve-locity Vector Potential Field-based Model Predictive Control (DVVPF-MPC) framework for autonomous vehicle path tracking. The proposed method integrates real-time local environmental perception with a dynam-ically guided potential field into the MPC structure, enhancing adaptability and robustness in trajectory tracking. By incorporating obstacle information and dynamic guidance into the potential field, the DVVPF-MPC significantly improves tracking accuracy and sys-tem stability in complex scenarios. Simulation results demonstrate that the proposed approach outperforms conventional MPC methods in terms of path-following precision and adaptability.