Dynamic Path Planning for Unmanned Surface Vehicles Based on Cooperative Air-Sea Coordination

This paper proposes an innovative air-sea collaborative dynamic path planning method for unmanned surface vehicles (USVs) exploring unknown waters, addressing the issues of missing global map information and low efficiency of traditional mapping methods. In scenarios lacking prior map information, existing solutions that rely on a single USV equipped with depth sensors for environmental perception face limitations such as mapping accuracy affected by wind and waves, and insufficient real-time performance. This study constructs an air-sea collaborative system using an unmanned aerial vehicle (UAV) to conduct aerial surveys of the target waters. A high-precision grid map is generated based on image binarization algorithms, and global path planning is completed using the A* algorithm. By integrating UAV GPS positioning data with camera azimuth parameters, the latitude and longitude coordinates of the navigation path are accurately calculated and transmitted in real-time to the USV for execution. At the same time, the USV is equipped with lidar to detect dynamic obstacles. When encountering an obstacle that the drone does not correctly identify, the USV automatically triggers the local path replanning mechanism. Experimental results show that the proposed method can effectively overcome the drawbacks of traditional navigation tasks that need to know the prior map in advance and electronic charts cannot accurately obtain real-time maritime map information through the coordinated operation of air and offshore platforms.