非平衡系统可靠度优化指派问题的混合求解算法  

Hybrid Genetic Algorithm for System Reliability Optimization of Unbalanced Component Assignment Problem

作　　者：刘琴 申海[1] 黎建强 LIU Qin;SHEN Hai;LAI Kinkeung(School of Business,Xi’an International Studies University,Xi’an 710128,China;School of Economics,ShenZhen University,Shenzhen 518060,China)

机构地区：[1]西安外国语大学商学院,陕西西安710128 [2]深圳大学经济学院,广东深圳518060

出　　处：《工业工程与管理》2020年第5期15-22,共8页Industrial Engineering and Management

基　　金：国家自然科学基金青年项目(71503197);陕西省教育厅科研计划项目资助(17JK0639);西安外国语大学科研计划项目(19XWD18)。

摘　　要：针对部件可互换系统的非平衡可靠度优化指派问题进行建模。模型以系统可靠度最大化为优化目标,考虑不同部件分派成本和总分派成本预算约束。求解算法首先采用基于Birnbaum重要度的局部搜索程序找到局部最优解,在此基础上根据指派问题的组合优化的特性设计混合遗传算法,该算法具有良好的收敛性和强大的搜索性能。通过小规模及大规模系统算例对算法的求解效果进行了验证。算例实验结果表明设计的混合算法能够提高求解质量,证明了算法的适用性和有效性。该算法为系统的可靠度优化设计研究提供了一条可借鉴的思路。Unbalanced Component assignment problem was proposed for systems with interchangeable components.The objective was to maximize the system reliabilty.A total assignment budget was considered as a constraint in the model. Local search procedure was embeded in genetic algorithm to find a local optimum and to improve the performance of the algorithm. A hybrid genetic algorithm was designed to solve this combinatorial optimization problem. The proposed algorithm has good convergency and search capability.Numerical experiment was conducted to verify the performance of the algorithm.The results of numerical experiment reveal that the proposed hybrid genetic algorithm can improve solution quality,which provides some implication to the system relibility optimization research.

关 键 词：最优分派 混合遗传算法 系统可靠度 

分 类 号：O224[理学—运筹学与控制论]