Risk Management of Supply Chain Networks with Dependent Disruptions

该奖项的研究目标是开发一个框架，以研究受依赖中断影响的供应链中的风险管理问题的关键问题。它为不确定的生产能力建立了一个多变量马尔可夫链模型，以捕捉中断的时间和空间影响。在具有多源和装配结构的两级供应链的背景下，本研究考察了供应链结构和相关中断对供应链绩效的影响。除了两级供应链研究中常用的工具外，本研究还运用了几个活跃/新的研究领域的创新方法，包括多变量依赖理论、Copula和Vine Copula，来研究具有依赖中断的供应链的定性行为和定量性质。交付成果包括具有相关中断/随机产量的新供应链模型、性能界限、评估系统性能的有效计算/模拟方法和最优订购策略。本研究是第一批使用冲天炉和藤条Copula方法研究供应链中的相依随机中断问题的研究之一。如果研究成功，本研究的结果将导致在中断中包含随机相依的供应链的最优设计和控制。本文的创新之处在于：(1)在战略层面，企业可以分析不同地理位置的供应节点的随机相关性及其对需求节点的影响；(2)在运营层面，它将帮助企业确定最优的库存策略、成本和客户服务水平。研究工作和成果将通过出版物和网站传播。商业和工程专业的研究生和本科生将从课堂教学和参与研究中受益。将开发一个软件包，使用户能够解决相关的供应链中断问题。

The research objective of this award is to develop a framework to study crucial issues of risk management problems in supply chains subject to dependent disruptions. It develops a multivariate Markov chain model for uncertain production capacities to capture temporal and spatial effects of disruptions. In the context of two-echelon supply chains with either the multi-source or assembly structure, this research investigates the effect of the supply chain structure and the dependent disruptions on supply chain performance. In addition to the commonly used tools in the study of two echelon supply chains, this research also utilizes innovative methodologies from several active/new research areas, including multivariate dependence theory, copula and vine copula to study the qualitative behavior and quantitative properties of supply chains with dependent disruptions. Deliverables include new supply chain models with correlated disruptions/random yields, performance bounds and efficient computational/simulation methods to evaluate system performances and optimal ordering strategies. This research is among the first ones to address dependent stochastic disruption problems in supply chains using cupola and vine copula methods.If successful, the result of this research will lead to optimal design and control of supply chains incorporating stochastic dependence in disruptions. The novelty of the proposed research lies in: (1)at the strategic level, firms can analyze the stochastic dependence of the supply nodes operating in different geographical locations and their impact on the demand nodes, and (2) at the operational level, it will help firms in determining optimal inventory policies, costs and customer service levels. Research efforts and results will be disseminated through publications and web sites. Graduate and undergraduate business and engineering students will benefit through classroom instruction and involvement in the research. A software package enabling users to solve dependent supply chain disruption problems will be developed.