基于数论佳点集的遗传算法初始种群均匀设计

文章利用数论中的佳点集理论和方法,给出了遗传算法初始种群生成的一种具有良好多样性的均匀分布设计.通过对遗传算法机理的研究,发现初始种群的分布状态不仅直接关系到遗传算法的全局收敛性,还影响算法的搜索效率,所以对初始种群进行科学合理设定是应用遗传算法进行寻优计算的一个重要问题.基于优化设计思想,提出应用佳点集均匀设计方法确定遗传算法的初始种群.这种方法具有简单易行、种群多样性好、更适合多维情况等特点,实验结果验证了该方法可以有效地改善算法的全局收敛性,提高搜索效率.

Uniform Design of Initial Population of Genetic Algorithm Based on Good Point Set

In this paper, on the basis of the good-point set theory,a method is proposed to establish initial population with good diversity by uniform design.By analyzing the genetic algorithm,a conclusion can be drawn that distribution of the initial population directly concerns global convergence and searching efficiency of genetic algorithm.The reasonable setting of initial population is an important problem in the application of genetic algorithm to performing optimization calculation. Based on optimization design theory,a good-point set method is proposed to establish initial population by uniform design. This new method has superiority in simplicity,diversity,multi-dimension, Simulation results show that this new method can improve global convergence and speed effectively.