A hybrid approach based on multi-agent geo simulation and reinforcement learning to solve a UAV patrolling problem
Jimmy Perron, Bernard Moulin, Jean Berger, Jimmy Hogan, Micheline Bélanger · 2008
In this paper we address a dynamic distributed patrolling problem where a team of autonomous unmanned aerial vehicles (UAVs) patrolling moving targets over a large area must coordinate. We propose a hybrid approach com-bining multi-agent geosimulation and reinforcement learn-ing enabling a group of agents to find near optimal solu-tions in realistic geo-referenced virtual environments. We present the COLMAS System which implements the pro-posed approach and show how a set of UAV can automati-cally find patrolling patterns in a dynamic environment characterized by unknown obstacles and moving targets. We also comment the value of the approach based on lim-ited computational results. 1