求整体优化全部解的区间排除遗传算法  

Appling interval exclusion genetic algorithms to finding all solutions of global optimization

作　　者：鞠训光[1] 于洪珍[1] 

机构地区：[1]中国矿业大学信息与电气工程学院,徐州221008

出　　处：《东南大学学报（自然科学版）》2006年第S1期82-85,共4页Journal of Southeast University：Natural Science Edition

基　　金：江苏省自然科学基金资助项目(BK2001073)

摘　　要：为了改善单纯遗传算法中的模式欺骗和早熟现象,应用区间算法和遗传算法的理论、方法构造了一种区间排除混合遗传算法,并采用二阶Lipschitz扩展作为检验极值的条件,用计算实例验证了本算法在一类整体优化问题中具有收敛到全部解的优良性能,检验条件简单、容易实现,明显提高了收敛的速度、可靠性,很好地解决了模式欺骗和早熟现象.To solve the mode deception and premature convergence problem, based on the theory and methoddogy of interval and genetic algorithms,interval exclusion genetic algorithms(IEGA) are constructed and quadratic extension of Lipschitz is applied to testify the interval extremum.The calculating examples validate the algorithm s excellent performance in the global optimization problem.The verifying terms are simpler and easy to be actualized.The algorithms propose can speed up the convergence with good reliability...

关 键 词：区间排除遗传算法 早熟现象 模式欺骗 

分 类 号：TB114.1[理学—运筹学与控制论]