Performance Study of Experienced Agent Joining Ad-Hoc UAV Teams

نویسندگان
1 Zanjan Electricity Power Distribution Company, Zanjan, Iran
2 Faculty of Engineering, University of Zanjan, Zanjan, Iran
3 Zanjan Regional Electric Company, Zanjan, Iran
چکیده
 In an autonomous robotic team, agents cooperate with each other to maximize the utility. The decision making and the manner of their operation and cooperation is a complex process due to dynamic environment, continues parameters, uncertain environment and unknown teammates. In this article, the persistent surveillance mission of unmanned aerial vehicles (UAVs) as a multi-agent system in the real world and the online decision making issue of agents are considered when the teammates and the environment are completely unknown. In this work, we propose a greedy algorithm to select an adaptive strategy based on the experiences it has before. Our experiments in a UAV persistent surveillance mission, show that this method improves the team performance in ad hoc teams. 

کلیدواژه‌ها