混合遗传算法在带走道的双目标布局问题中的应用  

The Application of Genetic Algorithm in the Bi-criteria Layout Problem with Aisles

作　　者：叶慕静[1] 周根贵[1] 

机构地区：[1]浙江工业大学经贸学院,浙江杭州310014

出　　处：《系统工程理论与实践》2005年第10期101-107,共7页Systems Engineering-Theory & Practice

基　　金：浙江省自然科学基金(No.Y104171)

摘　　要：根据系统布置设计思想,以物流成本和非物流关系密切度作为双目标,提出带有固定的纵、横走道的矩形平面布局模型.由于该布局问题属于NP-完全问题,构造一个带局部搜索的多目标遗传算法求解该问题的Pareto非劣解集.文中将两个不同量纲的目标函数量纲标准化,并以随机权重式评价目标函数.数值分析表明混合遗传算法对于中等规模该布局问题是十分有效的.According to the idea of Systematic Layout Planning, this paper proposes a bi-criteria mathematic model concerning the material handling costs and nonmaterial relation requirements for a layout problem with fixed lengthways and transverse passageways between sections. Since this problem is NP-hard, we develop a new multi-objective genetic algorithm approach hybridized with local search to obtain the set of its Pareto solutions. In this paper, we adapt a random weight approach to combine the normalized value of two objectives. The numerical experiments show the great effectiveness of the proposed method in dealing with layout problems in moderate scales.

关 键 词：布局问题 双目标 遗传算法 局部搜索 

分 类 号：TP138[自动化与计算机技术—控制理论与控制工程]