A navigation path planning algorithm for USVs in restricted inland waters based on the optimization and fusion of A* and DWA

Path planning with active obstacle avoidance is essential for navigation safety of unmanned surface vehicles (USVs) in restricted inland waters. To address insufficient safety, tortuous paths, and unstable tracking, this study proposes IA*D, integrating an improved A* algorithm with an improved Dynamic Window Approach (DWA). In global planning, a goal-oriented five-direction search strategy, a diagonal corner-cutting prevention constraint, and an adaptive risk-aware heuristic function improve search efficiency and path safety, while the Floyd algorithm removes redundant nodes to enhance smoothness. In local planning, LOS-guided detection dynamically selects the optimal local target point within the line-of-sight range, reducing oscillation and deadlock caused by discrete waypoint tracking. A smoothness factor is incorporated into the DWA evaluation function to suppress heading oscillation. Simulation results show that IA*D generates collision-free paths in static environments and improves obstacle avoidance, tracking stability, navigation efficiency, and safety under dynamic conditions.