Overtaking Path Planning for CAV Based on Improved Artificial Potential Field

Overtaking path planning is a key technology for Connected and Autonomous Vehicles (CAVs). However, existing path planning methods have limitations, such as safety and comfort. This study proposes a driving potential field model to improve the autonomy, safety, and comfort of CAVs during overtaking. The method builds an improved gravitational and repulsive potential field by considering the relative velocity and relative acceleration of the CAV and the leading vehicle. The influence distances of the forces are also specified to make the magnitude of the forces more specific. Road boundary potential fields are also established to regulate the driving area of the CAV. The overtaking path planning algorithm is tested by co-simulation of MATLAB/Simulink, CarSim, and PreScan based on four existing overtaking scenarios in the Chongqing Inner Ring Expressway. The results show that the proposed algorithm can significantly reduce the curvature of the overtaking path by 62.69%, 83.86%, 85.07%, and 91.18%, respectively, compared to the traditional artificial potential field method. The lateral acceleration is also significantly reduced by 36.33%, 62.97%, 45.77%, and 53.38%, respectively, in different overtaking scenarios. The lateral acceleration is less than 0.13 m/s
²
in all cases. Overall, the proposed algorithm is effective in improving overtaking path planning in terms of safety, efficiency, and comfort.