遗传算法局部最优的定向移民策略

Directional Immigrant Strategy for Local Optimization of Genetic Algorithm

基于定向移民策略的改进遗传算法,把定向移民引入遗传算法,经定向移民产生初始种群的集合,再复制、交叉和变异进化,随机选取高品质移民,组成新种群。由新种群开始复制、交叉和变异,直至寻出最优个体。试验结果表明,初始种群克服了GA自适应随机搜索局部最优和过早收敛问题。

The improved genetic algorithm (GA) based on directional immigrant strategy is that directional immigrant was inducted into GA. The set of original group was produced by directional immigrant, after copying, crossing, mutation evolution, and choose excellent immigrant to make up of new group. The new group begins to copy, cross and mutate until the best unit will be found. The experiment result shows that problems of local optimization and premature constringency were resolved with the original group in GA adaptive random search.