Using genetic algorithms to evolve the control rules of a swarm of UAVs

Due to the large number of interactions that the agents in a swarm of UAVs have with each other as well as with their environment, it is necessary to obtain a viable procedure that yields a reasonable group behavior from these local interactions. This paper proposes a hierarchical behavior-based model in which several parameters are adjusted with a genetic algorithm (GA). The presented model implements three explicit layers of behaviors (basic, group and mission) in a simulation in which the agents seek to survey a rectangular target area while avoiding a circular obstacle