基于协进化机制的Multi_agent分布式智能控制体系_罗杰.pdfVIP

32 19 2006 10 Vol.32 19 Computer Engineering October 2006 · · 2006 A TP27 Multi-agent 1,2 1 1 1. 100022 2. 330027 Agent Agent Architecture for Multi-agent Distributed Intelligent Control Based on Coevolution Mechanism 1 2 1 1 LUO Jie , DUAN Jianmin , CHEN Jianxin (1. College of Electronic Information Control Engineering, Beijing University of Technology, Beijing 100022; 2. College of Computer Information Engineering, Jiangxi Normal University, Nanchang 330027) Abstract This paper presents a coevolution-based multi-agent distributed intelligent control architecture. It adopts the method of decomposing the complex distributed control system into basic behavior modules which can be set regulable activity, and constructs them into multi-agent system. The coevolution-based distributed parallel learning structure and its coevolution algorithms are shown. It makes every basic behavior rule-base and global behavior coordination rule-base coevolve. This architecture is simple but effective, and its local control and global behavior coordination obtain adaptive online learning capability. The expansibility of this system is well. The system is practical, and may be applied to embedded control system operating at limited resource. Key words Distributed intelligent control; Multi-agent system(MAS); Coevolution; Coordination and cooperation; Architecture 1 Agent Agent