一类自适应遗传算法

New Adaptive Genetic Algorithm

简要介绍了徐宗本教授新近提出的可分解 /可拼接遗传算法编码 ,并证明了一个利用对偶适应函数判定已知个体是否为相对全局最优解的判据 .基于新的编码与判据 ,分别在种群层次与基因层次发展了动态变异与动态选择操作 ,进而提出了一类自适应遗传算法 .对其原理的分析表明 :新算法可用于求解所给问题的任意精度解 ,且在很大程度上可避免遗传算法早熟现象 .

Self adaption is one of the most promising areas of research in the evolutionary computation. The splicing/decomposible encoding scheme proposed by Xu et al is applied to the construction of dynamic selection and mutation operators that were built respectively on the levels of genotype and phonotype of population. A completely new adaptive genetic algorithm is thus developed. The analysis of the principle indicates that the new algorithm can obtain the arbitary resolution of the given problems and can avoid the premature convergence of the genetic algorithm in a great extent. It is a profitable research of the adaptive computation technology.