Rail-truck intermodal transportation plays a vital role in freight transportation in North America, and hence a crucial issue is to ensure continuity and minimize the adverse impacts from disruption. We propose a framework based on optimization and regression analysis for recovery from random disruptions of rail intermodal terminals. Two mixed-integer programming models are developed to simulate the normal and post-disruption operations, and the outputs are used in a predictive analytics model to identify the determinants of terminal criticality. The proposed framework is applied to a case study built using the realistic infrastructure of a rail-truck intermodal network in the United States. The resulting analyses underscore the importance of implementing the mitigation strategy at a marginal increase in pre-disruption cost, which in turn not only improves network resiliency but also results in significant cost savings by not using the more expensive recovery strategies. We further show that intermodal train service design resulting in a more balanced distribution of freight in the network could also help reduce terminal criticality, and that it makes sense to have backup rental capacities for some terminals.

铁路卡车多式联运在北美货运中发挥着至关重要的作用，因此关键问题是确保连续性并最大限度地减少中断带来的不利影响。我们提出了一个基于优化和回归分析的框架，用于从铁路联运码头的随机中断中恢复。开发了两个混合整数编程模型来模拟正常和中断后的操作，输出用于预测分析模型，以确定终端关键性的决定因素。所提出的框架应用于使用美国铁路-卡车多式联运网络的现实基础设施构建的案例研究。由此产生的分析强调了在中断前成本略有增加的情况下实施缓解策略的重要性，这反过来不仅提高了网络弹性，而且通过不使用更昂贵的恢复策略来显着节省成本。我们进一步表明，多式联运列车服务设计使网络中的货运分布更加均衡，也有助于降低终端的关键性，并且为某些终端提供备用租赁能力是有意义的。