基于协进化机制的Multi-agent分布式智能控制体系  

Architecture for Multi-agent Distributed Intelligent Control Based on Coevolution Mechanism

作　　者：罗杰[1] 段建民[1] 陈建新[1] 

机构地区：[1]北京工业大学电子信息与控制工程学院

出　　处：《计算机工程》2006年第19期34-37,共4页Computer Engineering

基　　金：教育部博士学科点专项科研基金资助项目;北京市自然科学基金资助重点项目

摘　　要：提出了一种基于协进化机制的多Agent分布式智能控制体系结构,采用划分基本行为模块的方法,将复杂、分布的控制系统构造成基于行为的、行为活性状态可调控的多Agent系统,并设计了基于协进化机制的分布式并行协进化学习机构及其协进化算法,使系统能分布并行地协进化各基本行为规则库和全局行为协调规则库。该体系结构能简单而有效地使系统的局部控制及全局行为协调协作具有较好的在线学习、自适应特性,系统可扩展性强,有较好的实用性,能应用于资源受限的嵌入式控制系统。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.

关 键 词：分布式智能控制 多智能体系统 协进化 协调协作 体系结构 

分 类 号：TP27[自动化与计算机技术—检测技术与自动化装置]