近年来，由于全球供应链的脆弱性不断加剧，提升供应链韧性已经成为学术界和行业实践者极为关注的议题。深入分析网络拓扑结构及其内部各利益主体间的相互作用变得尤为关键。本论文采用多种博弈论和优化方法，旨在从战略层面与运营层面构建更具韧性的供应链网络，以应对未来不确定性的风险。第2章着眼于供应链网络中存在供应中断风险的情境，研究了一个下游买方与多个竞争性上游供应商之间期权契约的设计。本研究将网络参与方之间的策略互动建模为斯塔克伯格博弈，并刻画了均衡条件下供应商的最优报价策略以及期权契约能够实现供应链网络全体成员双赢的条件，通过数值试验表明期权契约能够显著增强供应网络韧性。第3章探讨了在多层供应链网络中延期付款机制设计的有效性，主要研究下游企业是否应将付款决策委托给其直接上游供应商，而非直接选择延期付款合同。我们发现在高利润率情况下，委托付款可能提高供应链韧性；然而，在低利润率情况下，委托付款可能会削弱供应链韧性。我们还拓展了生产成本的有限可见性、零售价格的内生设定，以及多个二级供应商之间的竞争三种情况，以检验上述发现的鲁棒性。第4章提出一个网络枢纽设计模型，该模型将战略层面的位置选择和容量决策与操作层面的路径和枢纽分配决策相结合，以解决考虑拥堵和异质性规模经济的枢纽网络设计问题。我们还采用了新的基于流量的混合整数二阶锥规划建模，并利用真实数据集进行了数值实验。数值研究结果可以为物流公司设计运输网络时提供管理学建议。
Improving resilience has recently attracted enormous interest from academia and practitioners because of unavoidably increasing vulnerabilities in global supply chains. Consequently, it is imperative to consider the structure of the network topology and examine the interactions among various stakeholders. This dissertation presents three papers that explore different game theory and optimization approaches at both the strategic and operational levels to address the challenges in building the resilient supply chain.Chapter 2 study considers procurement with supply options in a supply chain network with one downstream procurement agency and multiple upstream suppliers with supply disruptions. We formulate the strategic interactions between the firms in the network as a Stackelberg game. We provide conditions under which supply options lead to “win-win” outcomes for all the firms in the supply chain network, and show that supply options significantly improve the resilience of the network. Chapter 3 examine the effectiveness and design of the deferred payment in a multi-tier supply chain network, to provide understanding of when (or whether or not) the most downstream firm should delegate the payment decision to its immediate upstream supplier, rather than directly select the deferred payment contract. We find that the delegation may improve resilience with a high margin but dampen the resilience with a low margin. We also show the robustness of these findings by extending the analysis to incorporate the limited visibility on the production cost, endogenously set retail price, and competition between multiple tier-2 suppliers.Chapter 4 propose a joint model that links the strategic level location and capacity decisions with the operational level routing and hub assignment decisions to solve hub network design problem with congestion and heterogeneous economics of scale. We also develop a novel flow-based mixed-integer second-order cone programming (MISOCP) formulation. We perform extensive numerical experiments on a real-world data set to validate the efficiency of solving the MISOCP reformulation. The numerical studies yield observations can be used as guidelines in the design of transportation network for a logistics company.
