Collaborative Path Planning Method for Heterogeneous Unmanned Swarm

This paper addresses challenges in collaborative path planning for heterogeneous unmanned swarms consisting of fixed-wing unmanned aerial vehicles (UAVs) and unmanned surface vessels (USVs) in both aerial and maritime environments. These challenges include slow map searching speeds, frequent path changes, and difficulties in maintaining inter-vehicle space. To overcome these issues, a two-layer collaborative path planning algorithm is proposed. The algorithm divides the path planning problem into two layers: lower-layer USV path planning and upper-layer UAV-to-USV following control. For the lower layer, USV path planning employs an improved A* algorithm that incorporates turn costs and cumulative cost decay factors. This modification reduces the number of nodes in the open set and accelerates pathfinding. Additionally, unnecessary turn points are eliminated by analyzing straight line reachability at turn points. Moreover, second-order Bezier curves are utilized to smooth paths, resulting in shorter distances, reduced turns, and smoother trajectories. In the upper layer, a hovering follow control algorithm compensates for disturbances in following distance due to relative motions between UAVs and USVs, thereby enhancing the precision of the following system. Through comparative analysis with standard algorithms, this study validates that the proposed lower-level pathfinding algorithm achieves faster speeds and smoother paths, while the upper-level tracking algorithm significantly reduces tracking error.