Multi-Agent协同进化算法研究  

Research on Multi-Agent Co-evolutionary Algorithm

作　　者：周铁军[1] 李阳[1] 

机构地区：[1]中南林业科技大学现代教育技术中心,长沙410004

出　　处：《计算机工程》2009年第13期205-207,共3页Computer Engineering

摘　　要：与传统优化方法相比,进化计算具有内在的并行性和自组织、自适应、自学习等智能特征,它在许多领域显示出巨大优势并取得一定成功。研究Multi-Agent协同进化算法,集成现有算法中的几种优势策略,利用混合策略的思想结合具体问题设计算法,并以实例说明该算法的有效性。In comparison with traditional optimization methods, evolutionary computation due to its intrinsic parallelism and some intelligent propertics, such as self-organizing, self-adaptation, and self-learning. Evolutionary computation has been successfully applied to many fields. This paper proposes a Multi-Agent co-evolutionary algorithm, designs a new hybrid algorithm by combining evolutionary algorithm with some field-special strategies, and proves their efficiency by several experiments.

关 键 词：多智能体 进化算法 蚁群算法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]