基于模块化和延迟生产的供应链建模与优化  

Modelling and Optimization of Supply Chain Based on Modularization and Postponement

作　　者：李院生[1] 叶飞帆[2] 方志梅[2] 杨建国[1] 

机构地区：[1]上海交通大学机械与动力工程学院,上海200030 [2]宁波大学工学院,宁波315211

出　　处：《中国机械工程》2004年第11期968-972,共5页China Mechanical Engineering

基　　金：国家自然科学基金资助项目 ( 70 1710 42 )

摘　　要：通过改善Ricardo模型 ,在供应链层次上建立四种供应链结构模型 ,利用各节点的成本和客户服务水平指标来衡量其性能。同时基于模糊偏好关系来确定各指标的权重 ,由此建立多目标优化模型。模型的求解利用Matlab中的遗传优化工具箱 (GAOT) ,并采用单因素方差分析验证了模型的鲁棒性。最后基于优化结果 ,分析了四种供应链结构的牛鞭效应。Based on the improvements of Ricardo model, a new model about four supply chain structures was set up on the supply-chain-level and then was evaluated by the criteria of cost and customer service. Meanwhile, based upon fuzzy preference, a multi-objective optimization model was established, by which the decision maker can decide the weights of different objectives. The model was optimized by the tool box GAOT in Matlab. In addition, the F-test results show the model has strong robustness. Finally, bullwhip effect on four supply chain structures was analyzed and compared on the basis of optimization results.

关 键 词：供应链 延迟生产 模块化 遗传算法 牛鞭效应 

分 类 号：TP14[自动化与计算机技术—控制理论与控制工程]