Path Tracking Control for Autonomous Vehicles Based on MPC Combined with Adaptive Potential Field Optimization

Path planning and trajectory tracking control are pivotal technologies for autonomous vehicles. The outcomes of path planning and trajectory tracking directly influence the stability and safety of vehicle operations. In this paper, we propose a method for establishing adaptive potential fields by considering the mass and speed of obstacles. This method is combined with model predictive control to enhance the safety and comfort of autonomous driving vehicles in complex scenarios. Initially, potential fields related to lane lines and road obstacles are established. Subsequently, by incorporating fuzzy rules, adaptive potential fields for obstacles with varying masses and speeds are designed. Based on vehicle dynamics model constraints and road constraints, these adaptive potential fields are integrated into the cost function of model predictive control, resulting in the design of a path planning controller algorithm based on adaptive potential fields. Finally, we validate the feasibility and robustness of the algorithm through an overtaking scenario. The results demonstrate that the designed algorithm can significantly improve driving safety and passenger comfort.