突发事件下基于贝叶斯更新的延迟生产策略  

Postponement production strategy under disruption based on Bayesian information update

作　　者：于辉[1,2] 王菲[1] 刘冠美 

机构地区：[1]重庆大学经济与工商管理学院,重庆400030 [2]重庆大学现代物流重庆市重点实验室,重庆400030

出　　处：《系统工程理论与实践》2015年第3期687-694,共8页Systems Engineering-Theory & Practice

基　　金：中央高校基本科研业务费(CDJKXB14004)

摘　　要：基于突发事件对多区域影响的不确定性导致应急物资生产策略和产量的决策困难,研究供应商如何选择生产策略进行应对的问题.构建了两时刻应急生产决策动态模型,并在第2时刻采用贝叶斯方法更新零售商遭遇突发事件概率的预测,从而给出两类生产策略:立刻生产和延迟生产.通过模型推导,对更新信息的价值以及两时刻损失差异的因素进行了定量化分析,进而给出供应商选择生产策略的准则及其最优决策产量.Abstract Uncertainty that incident affect many areas makes decision of emergency supplies production time and output difficult, this paper investigates how the supplier chooses the production strategy to deal with emergency. A dynamic two-moments emergency production decision model is established, by Bayesian method, the forecast probability of retailers encounter emergencies at the second moment is updated, based on this, two production strategies were proposed： immediately production and delay production. Through model deduction, the value of updating information and some key factors that reducing two-moment loss function different are quantitative analyzed. Furthermore, the production strategy selection criteria is proposed and the optimal production quantity is obtained.

关 键 词：突发事件 供应链 延迟生产 贝叶斯更新 

分 类 号：F273[经济管理—企业管理]