Using a genetic algorithm to develop rules to guide unmanned aerial vehicles

An unmanned aerial vehicle (UAV) is a remotely controlled plane with sensing devices that has the capability to fly over terrain in search of enemy activity. We investigate the use of a genetic algorithm to develop rules that guide the UAV by modeling the amount of uncertainty the UAV faces in terms of probability distributions over grid cells representing terrain. We employ the SAMUEL evolutionary learning system to create a set of rules with which to guide the UAV. Results indicate this methodology is capable of creating robust yet consistent sets of rules.