COLLECTIVE MOVEMENTS OF MOBILE ROBOTS WITH BEHAVIOR MODELS OF A FISH

This chapter presents a simulation model of multiple autonomous mobile robots with behavior models of a fish school. Although a school of fish does not need a special individual to lead it, an autonomous movement emerges from interactions among neighboring bodies. This study aims to realize autonomous collective movements of mobile robots through interactions among robots like a fish school. We used Khepera robots to simulate mobile cars running on a freeway. Autonomous behavior is assumed for the robots to run freely by sensing neighboring robots or the guard rails by means of infrared rays. An evaluation function was applied to free style running robots to run forward efficiently without colliding with other robots or guard rails. Genetic algorithms(GA) were applied to optimize the behavior models of a robot. As a result, we have obtained a nice collective movement of multirobots, with high techniques, such as speeding up, slowing down, and overtaking a slow robot ahead.