显隐性二倍体编码在MOGA中的应用及其可行性证明  

Application of Dominant-Recessive Diploid Codes in the MOGA and Its Feasibility

作　　者：李娜[1] 仁庆道尔吉[1,2] 曹军[3] 

机构地区：[1]内蒙古工业大学理学院,呼和浩特010051 [2]西安电子科技大学理学院,西安710071 [3]内蒙古大学数学科学学院,呼和浩特010021

出　　处：《内蒙古大学学报（自然科学版）》2010年第5期551-555,共5页Journal of Inner Mongolia University：Natural Science Edition

摘　　要：遗传算法是一种以种群进化为基础的全局优化搜索算法,并且自然界的高等生物的染色体结构形式是二倍体或多倍体,即有显隐性之分.结合以上两种理念,引进一种显隐性二倍体编码方法,并应用于多目标遗传算法(MOGA)中,通过分析在一般的锦标赛选择、两点交叉、基本位变异三种最基本的遗传操作作用下,模式在下一代的样本数,证明了显隐性二倍体编码的模式定理.从理论上说明这种编码在多目标遗传算法中是可行的,为进一步数值实验提供基础.Genetic algorithm is a global optimization search algorithm based on the population evolution, and the chromosome structure of advanced biology in nature is diploid or polyploid to show dominant or recessive. Combining these two concepts, a new dominative--recessive code method was introduced and applied in the multi-objective optimization. By analyzing the number of samples in the next generation, which was operated in three basic operations of tournament selection, two-point crossing,and the basic bit mutation,the Scheme Theorem is proved. It showed that such a code in a multi-objective genetic algorithm is theoretically possible to provide a foundation for the further nu- merical experiments.

关 键 词：显隐性二倍体编码 多目标遗传算法 解群 模式定理 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]