小生境人工免疫算法用于多峰函数优化  

An Artificial Immune Algorithm for Multimodal Optimization Based on Niche Techniques

作　　者：滕泓虬 李春华[2] 

机构地区：[1]广东科贸职业技术学院计算机系,广东广州510640 [2]华南理工大学自动化科学与工程学院,广东广州510640

出　　处：《计算机仿真》2009年第12期148-150,248,共4页Computer Simulation

基　　金：广东省工业攻关项目资助(710909)

摘　　要：基于自然免疫系统的原理,提出了小生境人工免疫算法,为了解决多峰优化问题,可使每个抗体适度得到修改。利用小生境的共享函数法更好地维持了群体的多样性,同时小生境淘汰算法和免疫系统的记忆及变异特性促使群体中优良个体持续向问题的优化解逼近。算法对三个典型的多峰问题进行了优化,给出了具体的实现步骤和较好的优化结果。An artificial immune algorithm based on niche techniques is applied to solve multimodal optimization problems. Not only can this algorithm well enhance the diversity of the evolutive population using a niche sharing technique, but also preserve elites of the population and continually evolve them as a result of a niche elimination technique and memory and mutation abilities of immune system. The algorithm is applied to solve three typical multimodal optimization problems and the results can perfectly map peaks of problems.

关 键 词：多峰优化 人工免疫算法 小生境 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]