摘要
该文主要针对遗传算法在实际操作应用中遗传算子和参数选取的复杂性,分析和探讨了一种多模式协作遗传算法。对于不同的优化问题以及遗传算法的不同运行阶段,引入局部遗传算法来实现一定领域中几种改进的遗传算子的适应性选取,使它们能够有效协作并且能够有效发挥各自特长,形成一种具有广泛适用性的算法模式。使用者可以在缺乏理论指导和经验参数情形下,不再凭借许多对比实验来选取合适的算子类型,方便和拓展了遗传算法的使用,并且通过试验证明了该方法的有效性。
In view of the complexity of operators＇ selection, this paper mainly analyses and describes a multi - pattern collaborative genetic algorithm. With regard to the different optimization problem and the different segment of genetic algorithm we use local genetic algorithm to realize the adaptive selection of several genetic operators in one field. The operators will collaborate effectively and develop their specialty sufficiently, then the algorithm pattern with extensive application will be formed. The user no longer depends on many contrast experiments to select appropriate operators and it is convenient for extending the use of genetic algorithms in the absence of theoretical instruction and experiential parameter. The results obtained with testing function are then presented to show the availability of the algorithm.
出处
《计算机仿真》
CSCD
2005年第12期136-138,190,共4页
Computer Simulation
基金
江西省第三批跨世纪学科带头人基金
江西省测控研究中心开放基金项目(2001012)
关键词
多模式协作遗传算法
局部遗传算法
遗传算子
适应性选取
函数优化
Multi - pattern collaborative genetic algorithm
Local genetic algorithm
Genetic operators
Adaptive selection
Function optimization
