Autonomous Vehicle Motion Planning

In dynamic and changing environments, ensuring the safe navigation of autonomous vehicles is still challenging. This paper proposes a motion planning method for autonomous vehicles that can navigate safely within everyday traffic environments. Considering vehicle safety, comfort, and operational efficiency, the method uses an improved artificial potential field approach for path planning, solving the optimal path within the S-L coordinate system. The introduction of an S-T graph is used for velocity planning, thereby transforming the velocity planning problem into a trajectory search issue. Dynamic programming is then employed to search for the velocity curve with the minimum cost value in the S-T graph, thus obtaining the optimal trajectory curve. The effectiveness of the proposed motion planning approach is verified through velocity planning experiments in scenarios involving following and overtaking other vehicles.