Abstract: To perform path planning for the unmanned surface vehicle (USV), which is one of the essential components of the USV's control system, we investigate a USV path-planning method based on the fast marching square (FMS). Firstly, we establish an evaluation function for the realistic marine environment of the USV to quantify the characteristics of the planned path under different Sats. Using the evaluation function, a suitable Sat is chosen to further conduct research on path planning in a dynamic environment. Appropriate improvements are made to the FMS method to optimize the path-planning process. When there is no collision risk, the FMS method is used to plan just one global path. When the USV detects other vessels, the FMS method is used for local path planning. In this process, when navigating along local paths, the velocity potential field is