基于多Agent系统的细菌群体趋药性算法研究  

Study on Bacterial Colony Chemotaxis Algorithm Based on Multi-Agent System

作　　者：华梁[1] HUA Liang （（1）ept. of Automatic Control, Suzhou Campus, Nanjing Institute of Railway Technology, Suzhou 215137, China）

机构地区：[1]南京铁道职业技术学院苏州校区自动控制系,江苏苏州215137

出　　处：《电脑知识与技术》2011年第3期1604-1606,1609,共4页Computer Knowledge and Technology

摘　　要：针对细菌群体趋药性（Bacterial Colony Chemotaxis，BCC）算法由于过度依赖群体交互而容易陷入局部最优解的缺陷，结合多Agent系统（Multi—Agent System，MAS）的主要特征构造一种全新算法——基于多Agent的细菌群体趋药性（MABCC）算法、该算法通过每个细菌Agent相互之间的竞争与协作，弱化其对群体信息的依赖，使其能够更精确地收敛到全局最优解．Bacterial Colony Chemotaxis （BCC）algorithm is easy to fall into a local optimal solution because of the over-reliance on group interaction. To solve the problem, this paper structured a new algorithm--Bacterial Colony Chemotaxis algorithm based on multi-agent （MABCC） combined with the main features of multi-agent system （MAS）. Through the competition and collaboration among the bacteria Agent, this algorithm weakens its dependence on the group information. As a result, the bacteria converge more accurately to the global （）ptimal solution. The simulation of some test function optimization using MABCC algorithm shows this algorithm is better than BCC algorithm for the global optimization.

关 键 词：细菌群体趋药性算法 多AGENT系统 竞争 协作 全局最优 

分 类 号：TP18[自动化与计算机技术—控制科学与工程;自动化与计算机技术—控制理论与控制工程]