Air-Ground Cooperation Optimal Coverage Path Planning with Connectivity Maintenance

Aerial-ground collaboration between unmanned aerial vehicles (UAVs) and unmanned ground vehicles (UGVs) can significantly enhance coverage efficiency, while it encounters challenges related to maintaining communication connectivity and optimizing multiple objectives. To address these challenges, this paper proposes a novel hybrid coverage path planning framework that integrates Multi-Agent Deep Deterministic Policy Gradient (MADDPG) and Particle Swarm Optimization (PSO), utilizing the strengths of MADDPG in handling complex multi-agent coordination and PSO in efficient global optimization to determine the optimal paths for UAVs and UGVs, thus enhancing coverage area. Specifically, MADDPG enables cooperative multi-UAV policy learning, optimizing UAV flight trajectories while ensuring stable connectivity with UGVs. Meanwhile, PSO optimizes UAV/UGV formation paths, keeping the UGVs within communication range and expanding the coverage area. Simulation results demonstrate that the proposed framework effectively maintains connectivity and maximizes coverage in dynamic environments.