基于数值计算方法的遗传算法的优化研究  

Optimization research of genetic algorithm based on numerical calculation method

作　　者：吴仕勇[1] 王天志[2] 接标[3] 李兴平[1] 

机构地区：[1]云南师范大学数学学院,云南昆明650092 [2]云南师范大学,云南昆明650092 [3]安徽师范大学数学与计算机科学学院,安徽芜湖241000

出　　处：《计算机工程与设计》2009年第12期2966-2967,3025,共3页Computer Engineering and Design

摘　　要：对其遗传算法的交叉及变异算子进行了深入研究,利用数值计算方法对遗传算法中的算子进行改进,提出了基于黄金分割法、最速下降法、Fibonacci法、"取大"、"取小"法的遗传算法的新算子。并通过实验对改进的遗传算法进行了实证检验,所得结果与传统算法进行了比较,改进算法行之有效,效果更加显著。The crossover operator and mutation operator of the genetic algorithm is studied, the genetic algorithms operator is improved by using of numerical computation methods, and the new operators is gotten based on golden section method, steepest descent method, Fibonacci method, minimal method. Genetic algorithms experiments on improved genetic algorithm are verified. The results of the improved algorithms are better than traditional algorithms which testifies the improved algorithms are notable.

关 键 词：数值计算方法 遗传算法 算子 黄金分割法 最速下降法 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]