基于置换的对偶遗传算法及其在动态排序优化问题中的应用  

Permutation-based dual genetic algorithm applied in dynamic sequencing optimizations

作　　者：刘黎黎[1] 汪定伟[1] 

机构地区：[1]东北大学信息科学与工程学院,沈阳110004

出　　处：《系统工程理论与实践》2008年第11期129-134,共6页Systems Engineering-Theory & Practice

基　　金：国家自然科学基金(70431003;70671020);创新群体项目(60521003);国家支撑计划项目(2006BAH02A09)

摘　　要：提出关于属性的对偶思想,将对偶机制的遗传算法推广到顺序编码动态优化问题中;引入群论中置换的概念表示对偶的属性,并提出基于置换的对偶遗传算法,该算法根据置换的性质构造学习算子和不完全对偶算子,在保持种群多样性的同时兼顾搜索精度.算法应用于交货期动态变化的单机总权重拖期时间最小化问题,取得了满意的结果.A new dual scheme with respect to the attribute is presented for extending the genetic algorithm with dual mechanism to solve order-encoding dynamic optimization problems; the conception of permutation in Group theory is borrowed to represent the attribute in the proposed permutation-based dual genetic algorithm. In order to maintain diversity as well as guarantee searching precision, a learning operator and an incomplete-dual operator are designed utilizing properties of permutation. It has been applied to the dynamic single machine total weighted tardiness problem with variational duedates, and satisfactory results have been achieved.

关 键 词：动态排序优化 遗传算法 置换 属性 不完全对偶 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]