食物链算法及其参数分析  

Analysis of Food-Chain Algorithm and Its Parameters

作　　者：喻海飞[1] 汪定伟[2] 

机构地区：[1]东北大学工商管理学院,辽宁沈阳110004 [2]东北大学信息科学与工程学院,辽宁沈阳110004

出　　处：《东北大学学报（自然科学版）》2007年第7期993-997,共5页Journal of Northeastern University(Natural Science)

基　　金：国家自然科学基金资助项目(704310037057107775103012)

摘　　要：根据人工生命突现集群与环境进行动态作用的特点,同时基于食物链也是生命系统中重要而又广泛存在的现象,通过定义各级人工生命的局部活动规则,提出一种具有食物链形式的人工生命算法,并称之为食物链算法.该算法是一类模拟自然生态系统进化和自然生物信息处理机理来进行优化计算的人工生命算法.阐述了基于人工生命的食物链算法的定义、思想和算法流程;并进一步讨论了食物链算法中的生命代谢能量规则,人工生命活动邻域的变化规则;对算法的关键参数进行了系统的分析研究.根据测试结果,本算法具有很强的类生命特性,比如自治、进化、自适应等,这些特性使得它非常适合于那些类生物系统优化问题的求解,如分销网络系统的选址分配问题.According to the characteristics of colony emergence of artificial organizes and the dynamic interaction between them and environment, the food-chain that is also an important and widespread phenomenon in the life system, and the rules of local activities of artificial organizes at different levels are defined to propose an artificial-life based algorithm referred as the food-chain algorithm. It is such a sort of artificial-life algorithms that optimize the computation by imitating the evolution of natural ecosystem and the information processing mechanism of natural organisms. Describes the definition, idea, and flow of the food-chain algorithm, and discusses the relevant rules of metabolic energy and change in neighborhood where the artificial-life individuals do something. Furthermore, the key parameters of the algorithm are systematically analyzed. The test results showed that the food-chain algorithm has such life-like traits as autonomous, evolutionary, and self-adaptive, which are highly fit for the optimization problems of life-like systems, such as the location-allocation problems of distribution network system.

关 键 词：人工生命 食物链算法 生态系统 突现集群 自适应行为 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] F270.7[自动化与计算机技术—控制科学与工程]