摘要
针对基本遗传算法难以实际应用的困难,吸收加速遗传算法的思想,提出一种考虑隔代遗传、模仿自然界中“附势”行为的广义遗传算法,它能够保持优秀个体的多样性,利用祖辈中的优秀个体变量变化空间作为下一代个体的繁殖空间.广义遗传算法概括了基本遗传算法和加速遗传算法,对它的参数进行不同设定时,可以设计出更多种类的遗传算法.将这种算法用于水流参数反演问题中,结果表明广义遗传算法收敛速度快,反演精度高,因此具有良好的应用前景.
Enlightened from Accelerating Genetic Algorithms (AGA), the author presented Generalized Genetic Algorithms(GGA) to settle the problem that Simple Genetic Algorithms(SGA) can't he easily applied to practice. GGA inherits ancestor's genes and imitates trend behavior in nature, It can preserve excellent individuals' diversity and uses excellent individual room of ancestors as propagating room of next generation. GGA generalizes SGA and AGA. When GGA' s parameters configured, more kinds of GAs may be designed. GGA was applied to identifying flow parameters, and the results indicated that GGA has rapid convergence rate and high convergence precision. Thus, GGA's future is hopeful.
出处
《系统工程理论与实践》
EI
CSCD
北大核心
2006年第2期133-137,共5页
Systems Engineering-Theory & Practice
基金
南京水利科学研究院开放流动研究基金(YK90501)
杭州市科技发展计划(20051331B06)
关键词
遗传算法
改进
广义遗传算法
水流参数
反演
genetic algorithms
improving
generalized genetic algorithms
flow parameters
identifying