Modeling and controlling of group behavior based on self-organizing principle

Abstract

This paper deals with a modeling and controlling strategy for multi-robot group behavior. While distributed autonomous robot system has advantages in flexibility of the system, it has also difficulties in coordinating the global system behavior as expected by the designer. Our research is to develop an approach in dealing with this problem. Self-organization is a natural principle which can be seen in the dissipative system. We present a hybrid top-down and bottom-up approach and a basic model for coordinating a macro scale group behavior based on the self-organizing mechanism. In particular, we realize the self-organization of temporal behavior patterns.