A Path Planning Method for Intelligent Ships Based on the Improved Artificial Potential Field Algorithm

Path planning for unmanned ships has become an important research topic in recent years. To enhance navigation safety and reduce collision risk, this study proposes an improved artificial potential field (IAPF) method. A route gravitational force is introduced to guide the ship back to the planned route after collision avoidance, while the repulsive force is optimized to improve path smoothness and obstacle-avoidance stability. A collision-risk-index-based repulsive force is further developed for dynamic obstacle avoidance, and its direction is modified according to the COLREGs. In addition, a time-sequential rolling path-planning framework integrating the APF and velocity obstacle algorithms is proposed to suppress path oscillation and adapt to target-ship maneuvers. The method is validated in head-on, crossing, and multiple-obstacle scenarios. The minimum passing distances are 1064.07 m, 1072.98 m, and 1109.66 m, respectively, and the maximum decision time is 148 ms. The results demonstrate that the proposed method can generate smooth, COLREGs-compliant paths, avoid local minima, adapt to dynamic encounters, and satisfy real-time collision-avoidance requirements.