遗传算法和自适应遗传算法的比较分析  

Comparison and Analysis on Genetic Algorithm and Adaptive Genetic Algorithm in Different Degrees of Fitness

作　　者：谢超凡[1] 徐鲁雄[1] 徐琳[1] XIEChaofan;XULuxiong;XU Lin(Fuqing Branch of Fujian Normal University,Fuqing Fujian 350300)

机构地区：[1]福建师范大学福清分校,福建福清35030

出　　处：《福建师大福清分校学报》2017年第5期22-26,共5页Journal of Fuqing Branch of Fujian Normal University

基　　金：福建省中青年教师教育科研项目(JAT170679)

摘　　要：相对于传统的优化算法,遗传算法在解决大规模非线性、非连续和复杂的综合性大型问题有着巨大的优势,易于计算机实现、操作简单、对问题域的超强描述能力,更为重要的是可实行并行计算.为了更贴近现实,自适应遗传算法在传统遗传算法的基础上进行了相应的调整,对遗传算法所涉及的参数进行自适应调整,因此具有更强的鲁棒性和更快的收敛能力.论文使用NK模型来生成不同程度的复杂度地形图,来比较分析对遗传算法和自适应遗传算法的性能影响,从而为实际问题对这两种算法的选择提供理论依据.Compared with the traditional optimization algorithm,the genetic algorithm has tremendous advantages in its strong description ability,simple operation in solving large-scale nonlinear and non continuous and comprehensive large-scale complexproblemsand its easy computer implementation of the problem domain.More important is the implementation of parallel computing.Adaptive genetic algorithm(GA)is extended on the basis of genetic algorithm,and the parameters of genetic algorithm are adjusted adaptively.This thesis employs the NK model to generate complex topography in different degree,to compare and analyze the impact on the performance of genetic algorithm and adaptive genetic algorithm,thus provides a theoretical basis for the practical problems of the two kinds of algorithm selection.

关 键 词：算法 遗传算法 NK模型 

分 类 号：O241[理学—计算数学]