一种基于多群体搜索的实数遗传算法

A Real Number GA Base on Groups Search

遗传算法在解决全局优化问题中起着重要的作用。本文依据部分动物物种在进化过程中的家族特征提出了一种多群体搜索实数遗传算法并且进行了仿真研究。通过仿真可以看,到该算法具有计算量较小,搜索精度高,收敛速度较快,抗早熟能力强,在有多个相同的全局最优点的情况下具有一定的搜索多个全局最优点参数的能力等特点。

Genetic algorithm plays an important role in solving global optimization problem. Based on peculiarity of some animal families, this paper proposes a real number Groups search GA. Simulation results show that this algorithm has merits of less computation, higher search veracity, rapider convergence, stronger ability against precocity, and can search multi global optimal points at the same time.