自动区域划分的分区域搜索狭义遗传算法

RESTRICTED GENETIC ALGORITHM OF AREA SEARCHING BASED ON AUTOMATIC AREA PARTING

用数学方法难以解决函数优化中的区域划分问题 .为此 ,提出了用狭义遗传算法实现区域划分的方法 .实现了基于自动区域划分的分区域搜索的狭义遗传算法 ,阐述了分区域搜索的控制策略 .从理论上分析了基于自动区域划分的分区域搜索的狭义遗传算法是全局收敛的 ,并具有收敛速度快、搜索过程稳定性高、可控制性强、便于并行实现等特点 .最后提出了有待研究的问题 .

Since it is difficult to solve the problem of area parting in function optimization by using mathematics methods, a method of area parting is put forward with restricted genetic algorithm in this paper. A restricted genetic algorithm of area searching, which is based on automatic area parting, has been implemented, and the control strategy of area searching is discussed also. While analyzed theoretically, the algorithm put forward in this paper is global convergence, and is provided with the following characteristics: higher convergence speed, higher stability, better controllability, and being implemented in parallel easily. Finally, some future researching problems are given.