采用不可微精确罚函数的约束优化演化算法  

Evolutionary Algorithm for Constrained Optimization Using Nondifferentiable Exact Penalty Functions

作　　者：周永华[1] 张旭 毛宗源[1] 

机构地区：[1]华南理工大学自动化科学与工程学院,广东广州510640 [2]齐齐哈尔第二机床厂工学院,黑龙江齐齐哈尔161000

出　　处：《小型微型计算机系统》2004年第8期1464-1467,共4页Journal of Chinese Computer Systems

基　　金：广东省自然科学基金项目资助 ( 0 1162 6)

摘　　要：针对多数已有的采用罚函数的约束优化遗传算法存在优化效果差的问题 ,提出了一种新的求解约束优化问题的演化算法 .借助不可微精确罚函数把约束问题转化为单个无约束问题来处理 .采用混合杂交和间歇变异来提高算法的搜索能力 .In view of that the performance of most of the previous genetic algorithms for constrained optimization using penalty functions is far from satisfactory, a new evolutionary algorithm for constrained optimization is proposed. Nondifferentiable exact penalty functions are used to transform a constrained optimization problem into a single unconstrained optimization problem. Hybrid crossovers and intermittent mutation are applied to increase the search capability of the algorithm. Numerical experimental results have shown the proposed algorithm outperforms most of the previous genetic algorithms using penalty functions.

关 键 词：演化算法 约束优化 不可微精确罚函数 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]