Path planning for autonomous vehicles under multi-road condition based on improved artificial potential field with A* algorithm

Abstract

Path planning for autonomous vehicles is a key technology in the field of intelligent transportation. In this paper, an improved Artificial Potential Field (APF) with A* algorithm is proposed for dynamic path planning of autonomous vehicles under multi-road conditions. Firstly, for the defects of APF algorithm, distance impact factor to the target and minimum gravitational value are respectively added for the APF potential field. Moreover, vehicle obstacle avoidance and lane-changing are realized by combining APF with Safety Distance Model (SDM-APF). Secondly, the heuristic function of the A* algorithm is improved to reduce redundant path points and enhance planning efficiency. The improved A* algorithm is integrated with the SDM-APF algorithm to further improve path planning performance. Finally, a traffic map of a multi-road scenario is designed for simulation using graphical user interface. The simulation results demonstrate that the improved A* algorithm outperforms the traditional approach, achieving a 2.5% reduction in redundant points and a 9.2% decrease in average intersection angles, along with a 2.0% increase in planning efficiency. The improved SDM-APF algorithm also demonstrates exceptional lane-centering capability and target accessibility, exhibiting almost 0 path volatility. From the initial point to the destination, even if the road conditions are varied and complex, the hybrid A*-SDM-APF algorithm can still successfully reach the destination, and ensure safety, smoothness, and stability of the vehicle trajectories. The simulation results provide strong evidence for the effectiveness and superiority of the algorithmic improvements and combinations.