Optimization of group behavior on cellular robotic system in dynamic environment

This paper addresses an optimization method of the group behavior on decentralized autonomous robotic systems. Decentralized autonomous robotic systems refer to multiple robotic systems, such as cellular robotic system (CEBOT). The CEBOT, which has been proposed by the authors, consists of a number of robotic units called cells. In the research on the CEBOT, the same as decentralized robotic systems, the optimization of the behavior of each robot is one of the most important issues, since it is difficult to predict the behavior of the each robot in multiple robotic systems. In order to optimize the behavior in a dynamic environment, the authors propose a concept of self-recognition for the decision making of the behavior in a group. According to the proposed concept, this paper shows the learning and adaptation strategy for the group behavior, and presents some simulation results