基于多目标模糊决策的混流装配线排序算法  

New algorithm for mixed model assembly lines sequencing problems based on Multi-objective fuzzy decision-making

作　　者：杨斐[1] 阚树林[1] 钱峰[1] 许洁[1] 张赋杰[1] 王越[1] 

机构地区：[1]上海大学机电工程及自动化学院,上海200083

出　　处：《机械设计与制造》2007年第12期209-211,共3页Machinery Design & Manufacture

基　　金：上海市教育委员会自然科学基金项目(编号:04AB46);上海市重点学科建设项目(编号:Y0102)

摘　　要：在订单生产环境下的多品种混合装配线排序问题的基础上,建立了三种目标函数:最小化生产周期时间、最小化工作站闲置与超载时间以及最大化平均负荷率。引入了一种基于模糊决策模型的改进后的混合遗传禁忌搜索算法。在求解过程中,计算出所有个体的各个目标特征值的影响度、影响度因子以及相对隶属度,将目标特征值矩阵转化为相对隶属度矩阵,应用模糊层次分析法,计算多目标排序问题中各指标的权重值,用综合评判线性加权平均模型进行评价,构造出相对适应度函数。On the basis of the sequencing problems of multi-product mixed model assembly lines for the order-based manufacturing, three objective functions were proposed： minimizing flow line tempo, minimizing total over loading and idle time and maximizing average load rate. An improved hybrid genetic-tabu search algorithm based on fuzzy decision-making model was presented. While seeking the answer to multi-objective decision-making model, the influence degree, the influence degree factor and the relative membership degree of the objective eigenvalue of each chromosome were computed. The objective eigenvalue matrix was changed into the relative membership degree matrix in each generation. The weight values of each indices in multi-objective sequencing problems were calculated with the theory of FAHP. Then, they were evaluated by the integrative judging linearity weighted model. The relative fitness function was set up.

关 键 词：遗传算法 禁忌搜索 混流装配线 多目标模糊决策 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]