摘要
在现有几种遗传算法(GA)的基础上,对GA中的适应度函数、交叉策略和变异策略做了进一步的设计,从而提出了一种新型改进GA。新型改进GA以群体的多样性与算法的收敛速度、全局与局部搜索能力的综合均衡为设计重点,较好地解决了一般GA收敛速度慢和局部搜索能力差的缺点。仿真结果表明:该算法与常用的标准GA和采用算术交叉算子的实值编码改进GA相比,有更快的收敛速度,更高的收敛精度及全局收敛概率。
Based on several existing Genetic Algorithm (GA), a improved GA is proposed through further facilitating the fitness function, crossover strategy and mutation strategy of GA. The new GA can effectively improve the poor performance of the existing one on convergence speed and local search ability because it emphasize the proportion of population diversity and convergence speed, and also the proportion of globe search ability and local search ability. Simulation results show that the new GA has faster convergence speed, higher accuracy and higher globe convergence probability compared with the standard GA and the arithmetic crossover based real -coded GA.
出处
《中山大学学报(自然科学版)》
CAS
CSCD
北大核心
2006年第2期36-40,共5页
Acta Scientiarum Naturalium Universitatis Sunyatseni
基金
国家自然科学基金资助项目(60575006)
关键词
遗传算法
适应度函数
交叉策略
变异策略
genetic algorithm
fitness function
crossover strategy
mutation strategy