Path Planning with Improved Curve Boundary Generation for Unmanned Surface Vehicle

This paper proposes a search-optimization-based path planning strategy for Unmanned Surface Vehicle (USV) is proposed by improving the boundary generation method. Compared with the existing sampling or graph theory approaches, both the accuracy and the efficiency of collision detection are considered in the proposed strategy. The path parameterization method provided in this paper can obtain a smoother continuous path within a reasonable time-consuming range, making it more compatible with speed or trajectory planning. Finally, numerical simulations are performed to verify the effectiveness of the proposed algorithm.