Reliability certification of supply chain networks under uncertain failures and demand
提出一个优化理论框架，通过灵活路由策略应对故障和需求不确定性，为供应链网络提供可靠性认证，帮助设计者评估并提升网络韧性。
Abstract In the dynamic and uncertain landscape of modern supply chains, a significant challenge for supply chain network (SCN) architects is ensuring that their designs consistently deliver reliable performance amidst variable traffic demands and unpredictable failures. This assurance is particularly difficult due to the exponential and potentially uncountable set of scenarios that must be considered for evaluation. In response to this challenge, we propose an optimization-theoretic framework to assess the performance of supply chain networks by incorporating flexible routing strategies amid system uncertainties. By modeling network failures and uncertain demand requirements, our methodology tackles the inherent intractability of these certification problems and provides valid upper bounds for the link utilization of the most congested link and the total unmet demand, essential metrics for maintaining reliable network operations. We leverage optimization and convexification techniques to optimize the SCN design under disruptions and demand variation. Through rigorous evaluations on various supply chain network topologies, our results demonstrate the efficacy and promise of our approach. Our framework not only provides a systematic way to certify SCN performance but also serves as a tool for supply chain architects to enhance resilience against unforeseen disruptions. Ultimately, our research contributes to the development of more reliable and resilient supply chains in an increasingly volatile global environment.