摘要
提出了一种新的遗传算法最优保存策略,该策略在最优个体保留的基础上,添加一个与最优个体相异因子较大,而适应值不过小的个体.这样做既利用了最优保存策略的全局收敛性,又通过新添加的个体来保持种群的多样性,以防止早熟现象的出现.对典型优化函数进行了测试,结果表明基于新的最优保存策略的遗传算法(DESGA)收敛性能显著好于一般简单遗传算法(SGA)和最优保存简单遗传算法(ESGA).该策略与最优保存策略一样具有一般通用性.
A new elitist strategy in genetic algorithms is presented, which reserves an individual in generation while reserving the optimal individual. The new lead-in individual has big dissimilarity factor to the optimal individual, and its fitness is not too small. In the new strategy, optimal individual is reserved to ensure globe convergence, and population diversity is preserved to prevent premature by the new individual. Simulation results to several typical function optimization problems show that the simple genetic algorithms with new strategy （DESGA） are more distinct than the simple genetic algorithms（SGA） and the simple genetic algorithms with the optimal individual reserved （ESGA）. As the elitist strategy, the new strategy has general versatility, and can be used in many genetic algorithms.
出处
《浙江大学学报（理学版）》
CAS
CSCD
北大核心
2006年第1期32-35,共4页
Journal of Zhejiang University（Science Edition）
基金
浙江省自然科学基金资助项目(197047)
关键词
遗传算法
最优保存策略
海明距离
相异因子
互补个体
genetic algorithms
elitist strategy
Hamming distance
dissimilarity factor
complementarity individual
