2A1-G16 人間の社会行動原理に基づく適応的協調理論 : 心理モデルを有するエージェントの協調(複数ロボットの協調制御)

概要

論文の詳細を見る
This paper addresses a novel concept for multiagent cooperation based on behavioral principle of human sociality. By adopted the human cooperation principles, multiple robots are expected to work more flexible and adaptive on environmental condition as human society. Since the optimal cooperation indexes could be variable depending on the environment, solution of collision and competition during the work are important problems to achieve high performance. The behavioral principle of human sociality is focused on because it has been achieved with human evolution and it seems adaptive to variable environments. In social psychological field, principles of human social behavior are investigated and social parameters are denned as the agent characteristics. Furthermore, we consider mental effects on each agent by changing dynamics parameters of agent motion. Then, the social principles, as leadership, group synchronization and recognition errors, are applied to the multiagent systems to improve the collision and competition problems. Finally, validity of the technique is verified by simulations and we found the mental effects improve the performance.
一般社団法人日本機械学会の論文
2008-06-06