自调节遗传算法的研究  

Study on Self-adjusting Genetic Algorithm

作　　者：庄健[1] 王孙安[1] 

机构地区：[1]西安交通大学机械工程学院,西安710049

出　　处：《系统仿真学报》2003年第2期281-282,286,共3页Journal of System Simulation

基　　金：陕西省自然科学研究基金资助(2001X17)

摘　　要：通过对普通遗传算法选择、交叉、变异三个算子的深入分析,揭示了早熟现象产生的原因,并数学证明了选择算子是早熟现象的主要原因。在此基础上设计了一种变异概率可以自调节的基因移民遗传算法,提出了群体相异度指标来衡量多样性,通过相异度自动地调节变异概率的大小,确保群体的多样性来避免早熟现象的出现。最后,应用马尔可夫链数学工具从理论上证明了该算法是收敛的。实验表明:该算法具有一定的通用性,适合于实际的工程计算。The selection operator, the corssover operator and the mutation operator of the genetic algorithm are deeply analyzed in the paper. Then, it is proved that the main reason of the premature phenomena is the selection operator. Based on the analysis, a self-adjusting gene migration genetic algorithm with the self-changeable mutation rate is designed. And the guideline of the population differentia is put forward. The probability of mutation is automatically changed by the population differentia. So the population diversity is kept and the premature phenomena are avoided. At last, the convergence of the algorithm is proved by the Markov-Chain. The experimental results indicate that it has the ability of the universal using and is suitable for the engineering calculation.

关 键 词：自调节遗传算法 MARKOV链 早熟现象 基因移民 

分 类 号：O242.23[理学—计算数学]