随机选址－库存配送网络集成规划模型及算法.PDFVIP

随机选址－库存配送网络集成规划模型及算法 1，3 2 1 秦绪伟 ，范玉顺 ,尹朝万 （1.中国科学院沈阳自动化研究所，辽宁沈阳 110016； 2. 清华大学自动化系，北京 100084; 3. 中国科学院研究生院，北京 100086） 摘要：研究了随机需求条件下由单供应商、候选分拨中心和分销点构成的选址－库存问题。综合考虑库存成本、 运输成本和设施成本之间的均衡关系，分销点、分拨中心分别基于周期检查（R, s, Q）和连续检查（s ,S）库存控制策 略，建立了二级库存与无能力约束选址集成规划模型。给出了适合求解实际规模问题的拉格朗日松弛算法，提出了求解 子问题的有效启发式方法，改进了次梯度优化方法。通过仿真试验验证了模型的正确性和算法的有效性。最后讨论了相 对于传统规划方法，需求方差、服务水平、持有成本、提前期等关键库存控制参数对系统运营成本节约的影响规律。 关键词：选址－库存问题；拉格朗日松弛算法 中图分类号：TP18;C934 文献标识码：A Stochastic location-inventory distribution network integrated design model and its Solution Algorithm 1，3 2 1 QIN Xu-wei ,FAN Yu-shun , YIN Chao-wan (1.Shenyang Institute of Automation , Chinese Academy of Sciences, Shenyang 110016, China, China; 2. Dep. Of Automation, Tsinghua Univ., Beijing 100084 3. Graduate School of the Chinese Academy of Sciences, Beijing 100086) Abstract: The Stochastic location-inventory distribution network design problem involving one supplier, candidate distribution centers and multiple retailers. Retailers adopt the periodic review （R, s, Q ）inventory policy, while distribution centers adopt the continuous review （s, S ）inventory policy. Considering the trade-off between inventory cost, shipping cost and facility location cost, the integrated the 2-echelon inventory cost function with uncapacitated facility location model is presented. A Lagrangian relaxations solution algorithm for realistic scale problem is proposed. A number of heuristics are outlined for solving effectively sub-problems. Sub-gradient method is improved. Simulations are given t