Study on intelligent collision avoidance and recovery path planning system for the waterjet-propelled unmanned surface vehicle

To avoid collision accidents and achieve independent recovery in the cluttered marine environment, this work proposes an intelligent path planning system for a waterjet-propelled unmanned surface vehicle (USV). Unlike the existing works on USV navigation systems, our study focuses on the real-time, smoothness and seaworthiness properties of the path in practical clutter environments. A hybrid A* algorithm with motion primitive constraints is proposed to generate an initial reference path. According to different types of dynamic obstacles, the International Regulations for Preventing Collisions at Sea (COLREGs) rules are used, and a local threat map based on the Apollonius circle is constructed to avoid these obstacles. Together with the motion characteristics of the waterjet-propelled USV, the Reeds-Shepp curve is used to calculate the autonomous recovery path. The performance of the proposed intelligent navigation system is verified in a practical marine environment. The experimental results show that our hybrid A* algorithm outperforms the conventional A* algorithm, and it is easy for the waterjet-propelled USV to follow the final continuous curvature path. It was demonstrated that our method can effectively avoid various static and dynamic obstacles in real time and thus help achieve autonomous recovery.