Robust and Stable Collision Avoidance for Unmanned Surface Vehicles

In order to assure the navigation safety of unmanned surface vehicles (USVs), this paper proposed a robust and stable collision avoidance method, which considers of obstacles in path and International Regulations for Preventing Collisions at Sea (COLREGS). Firstly, a priority sampling path planning method is proposed for complying COLREGS. Secondly, a kinodynamic path planning is carried out to improve success rate of path planning. At last, Physical constrains such as distance from obstacles, navigation distance, and heading change which are formulated as mathematical formula, are used for optimized object functions to get a safe, stable, and dynamically feasible route. The proposed method is tested extensively through typical scenarios which complied COLREGS rules, multiple random static and dynamic obstacles scenarios, and actual ship navigation scenarios.