管理评论

管理评论 ›› 2023, Vol. 35 ›› Issue (6): 275-287.

• 运作管理 • 上一篇    下一篇

服务水平约束下基于鲁棒优化的共享平台库存水平设置及分配策略

邱若臻1, 戢媛媛1, 邵松世2   

  1. 1. 东北大学工商管理学院, 沈阳 110169;
    2. 海军工程大学舰船与海洋学院, 武汉 430033
  • 收稿日期:2021-08-12 发布日期:2023-07-27
  • 作者简介:邱若臻,东北大学工商管理学院教授,博士生导师,博士;戢媛媛,东北大学工商管理学院博士研究生;邵松世,海军工程大学舰船与海洋学院副教授,博士。
  • 基金资助:
    国家自然科学基金项目(72272030;71772035)；教育部人文社会科学研究项目（22YJA630064）；中央高校基本科研业务专项资金资助（N2306007）。

Robust Optimization-based Inventory Level and Allocation Strategy for Sharing Platform under Service Level Constraint

Qiu Ruozhen1, Ji Yuanyuan1, Shao Songshi2   

  1. 1. School of Business Administration, Northeastern University, Shenyang 110169;
    2. College of Naval Architecture and Ocean Engineering, Naval University of Engineering, Wuhan 430033
  • Received:2021-08-12 Published:2023-07-27

PDF (PC)

113

摘要： 考虑由单个库存共享平台和多个零售商组成的供应链系统，研究了需求信息共享与不共享两种情况下的共享平台最优库存水平设置与库存分配问题，建立了仅知需求均值和方差信息下，考虑零售商服务水平约束的库存鲁棒优化模型。在需求信息共享下，库存共享平台可准确地获知零售商的需求均值和方差信息，将零售商服务水平约束转化为易于处理的机会约束问题，推导了最优库存水平和零售商库存分配量决策，分析了相关决策随系统参数的变化情况。在需求信息不共享下，库存共享平台仅知零售商需求均值和方差所隶属的不确定集，特别地，采用椭球不确定集刻画不确定需求均值和方差，运用对偶理论将含有机会约束的库存优化模型转化为易于求解的数学规划问题。最后，通过数值算例对比分析了信息共享对最优库存水平和库存分配策略的影响。结果表明，在不同信息条件下，共享平台库存水平均随需求均值、方差与服务水平要求的增加而增加；零售商共享信息对共享平台的库存优化有积极影响，零售商共享的信息越完善，共享平台持有的库存水平越低；当零售商不共享信息时，共享平台库存水平随其需求均值和方差的不确定性程度增加而增加，并且需求均值的不确定性对库存水平影响更大。

关键词: 共享平台, 库存分配, 服务水平约束, 鲁棒优化, 信息共享

Abstract: Assuming there is a supply chain composed of a single inventory sharing platform and multiple retailers, this paper studies the optimal inventory level and allocation strategies of the inventory sharing platform with and without demand information sharing. A robust optimization model is established, with the retailer's service level constraint taken into consideration and only demand mean and variance information known. Under the condition of information sharing, the inventory sharing platform can accurately obtain the retailers' demand mean and variance. The retailer's service level constraint is transformed into an easy-to-handle chance constraint problem. Under that condition, the optimal inventory level and allocation strategies are deduced, and the changes of the strategies with the changes of the related parameters are analyzed. Under the condition of not sharing demand information, the inventory sharing platform only knows the uncertain set of the retailers' demand mean and variance. In particular, the ellipsoidal uncertain set is used to describe the mean and variance of uncertain demand, the dual theory is used to transform the inventory optimization model and studies are conducted to analyze the impact of information sharing on the optimal inventory level and allocation strategy. The results show that, under different information conditions, the inventory level of the inventory sharing platform increases with the increase of the mean, variance and service level requirements; the information shared by the retailers has a positive impact on the inventory optimization of the inventory sharing platform, namely, the better information sharing among the retailers, the lower inventory level on the inventory sharing platform; when the retailers do not share information, the inventory level of the inventory sharing platform increases with the uncertainty of mean and variance increases, and the uncertainty of mean has a greater impact on the inventory level.

Key words: sharing platform, inventory allocation strategy, service level constraint, robust optimization, information sharing