摘要
遗传算法(GA)是利用自然选择和进化思想在高维空间中寻优的方法,其寻优过程始终保持整个种群的进化.本文提出了实数编码最优子种群遗传算法理论,通过从种群中选出适应值最高的若干数量的个体,组成该代最优子种群,将最优子种群中的个体与种群中其它个体进行交叉变异、最优子种群中的个体间也进行交叉变异,从而产生新的种群.该遗传算法使得遗传过程中落入局部最优解的几乎不可能,对于多极值问题非常有效,收敛速度也非常快.
Genetic algorithm （GA） is a method of using the idea of natural selection and evolution in higher dimensional space. Its optimal process always keeps all population evolving. In this paper a kind of new decimal genetic algorithm of optimal filial-population is put forward. A certain number of individual with high fitness are chosen from the population and made up an optimal filial-population, the individual of optimal filial-population are made crossover and mutation with the individual of population and the individual of population are made crossover and mutation each other, then the next new population is produced. This genetic algorithm makes it almost impossible to fall into local optimum, and it is very effective on muhi-extremum programming, its convergence velocity is very fast.
出处
《淮阴师范学院学报（自然科学版）》
CAS
2008年第2期101-104,共4页
Journal of Huaiyin Teachers College;Natural Science Edition
关键词
遗传算法
最优子种群
实数编码
genetic algorithm
optimal filial-population
decimal code
