单亲遗传算法的机理分析  

On Operating Mechanism of Parthenogenetic Algorithm

作　　者：李茂军[1] 罗安[2] 

机构地区：[1]长沙理工大学电气与信息工程学院,湖南长沙410076 [2]湖南大学电气与信息工程学院,湖南长沙410072

出　　处：《长沙理工大学学报（自然科学版）》2004年第1期76-79,共4页Journal of Changsha University of Science and Technology:Natural Science

基　　金：国家教育部博士点专项科研基金资助项目(20030533014)

摘　　要：单亲遗传算法(PGA)取消了传统遗传算法(TGA)的交叉算子,代之以仅在1条染色体上操作的基因重组等遗传算子,简化了遗传操作,提高了计算效率,并且不要求初始群体的多样性,也不存在"早熟收敛"问题.由于PGA采用单亲繁殖方式,不像TGA那样模拟自然界绝大部分生物的双亲繁殖方式,使得不少学者对PGA是否属于遗传算法的范畴表示怀疑,对PGA的运行机理也缺乏认识.本研究拟基于抽象系统的基本概念,构造PGA和TGA2个抽象系统,证明这2个抽象系统是同构的,从而证明了PGA和TGA是等价的.这就说明PGA和TGA具有相同的运行机理,也说明PGA和TGA一样是属于遗传算法的范畴.Partheno-genetic algorithm (PGA) repeals crossover operators of traditional genetic algorithms (TGA) while it has the reproduction manner with only parent. Comparing with TGA, the genetic operation of PGA is simpler; the initial population of PGA needs not be varied; there is no “immature convergence” in using PGA and the searching efficiency of PGA is higher. As PGA uses the reproduction manner with only parent abandoning the reproduction manner with double parents as most biobodies, most of the scholars think that PGA does not belong to the category of genetic algorithms. Based on concept of abstract system, two abstract systems, PGA and TGA are constructed and it is proved that the two abstract systems are isomorphic, so PGA has the same operating mechanism as TGA, that is to say, PGA belongs to the category of genetic algorithms as TGA.

关 键 词：单亲遗传算法 PGA 遗传算子 编码方式 抽象系统 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]