Path Tracking and Local Obstacle Avoidance for Automated Vehicle Based on Improved Artificial Potential Field

This study proposes an improved artificial potential field (APF) by considering the cooperative control of local obstacle avoidance and path tracking for automated vehicles. We established the path gravitational potential field (GPF) based on the scheduled path (SP), including the lateral and longitudinal GPFs, to enable the automated vehicle to quickly return to the SP and track after obstacle avoidance, while maintaining control of speed for the entire process. To address the local optimal solution problem of the classical APF, we proposed a sub-target-point selection strategy based on the information of obstacles and SP and established the GPF of the sub-target points. Thus, the automated vehicle can avoid obstacles and quickly return to the SP. Furthermore, the relative velocity of the automated vehicle and the obstacle was used to establish the velocity repulsion potential field (RPF), which improved the adaptability of the APF to dynamic obstacles. The simulation results indicate that the improved APF is capable of cooperative control of path tracking and local obstacle avoidance. Code is available at https://github.com/xiaowang617/Improve-APF .