摘要
分析了以往的遗传算法适应度函数设计通常只针对目标函数,而没有考虑自变量.将物种的概念引入遗传算法,提出了根据种子到当前最优点的距离将种群分为两个物种,一个为当前最优物种,另一个为物种仓库,对此两个物种分别以不同的交叉概率和变异概率进行遗传运算,用以平衡种群的"选择压力"和"种群多样性".数值结果表明了本方法的有效性和稳定性.
This paper point out the design of fitness function prevenient only based on the objective function, not taking account of independent variables of the problem. This paper introduces species into genetic algorithms, dividing the population into two species by the distances between individuals and optimal individual. One is the optimal species of current generation, and the other is a species warehouse. We apply genetic algorithm to two species with different crossover Probability and mutation Probability. Some numerical tests have been made and the results show that the algorithm is effective.
出处
《小型微型计算机系统》
CSCD
北大核心
2009年第3期534-536,共3页
Journal of Chinese Computer Systems
基金
国家重点基础研究发展规划“九七三”项目(2002CB312200)资助
关键词
遗传算法
物种
genetic algorithms
species