Multi-UAV Path Coordination through Generalized Potential Games

In this paper, the problem to safely and efficiently navigate multiple Unmanned Aerial Vehicles (UAVs) in complex environments is addressed. Each UAV aims at achieving an approximate optimal trajectory plan while considering constraints from the environment and other neighboring UAVs. The problem is formulated as a generalized potential game. An algorithm, following an iterative best response approach, is proposed. Experimental results verify the coordination of a team of UAVs for a given team goal description. Also, the time cost of computation for the proposed algorithm is investigated. The results show a high-level goal to navigate a team of UAVs can be automatically decomposed to safe and efficient trajectory plans for each UAV in the team by solving generalized potential games in a decentralized manner.