The exclusive and excessive use of long-distance road transportation is not suitable way to reduce the negative environmental impacts of logistics systems. Intermodal transport, combining road with other transport modes, has the potential to reduce both operating costs and carbon footprints. One of the reasons for the low share of intermodal transport is its requirement for the coordination of scheduled transport services that can result in reducing reliability in case of disruptions due to the arrival of new shipment orders, fluctuations in shipment quantities, delays, and service cancellations within the network. This calls for reliable and efficient algorithms to replan the shipments’ distribution. In this paper, the replanning problem is formulated as a path-based multi-commodity network flows. We provide two different network topologies, one of which is based on a time–space network, while the other embeds time aspect in a highly scalable alternative structure to large transportation networks. We propose a column generation method whose pricing sub-problems are presented as resource constrained shortest path problem solved via a tailored label-correcting algorithm. We look at the pros and cons of complete and partial replanning in case of disruption and provide managerial insights for intermodal networks. An extensive set of computational experiments is presented on realistic instances being generated with the consultation of our industrial partners for a logistic network including railways, waterways, and roads. The promising outcomes validate the efficiency of the proposed approach that can be easily adjusted to real-time intermodal logistic replanning.

排他性和过度使用长途公路运输不是减少物流系统对环境的负面影响的合适方法。多式联运将公路与其他运输方式相结合，有可能降低运营成本和碳足迹。多式联运份额低的原因之一是它需要协调定期运输服务，这可能会导致可靠性降低，以防由于新订单的到达、装运数量的波动、延误和服务取消而造成的中断网络内。这需要可靠且高效的算法来重新规划货运分布。在本文中，重新规划问题被表述为基于路径的多商品网络流。我们提供两种不同的网络拓扑结构，其中一个基于时空网络，而另一个将时间方面嵌入到大型交通网络的高度可扩展的替代结构中。我们提出了一种列生成方法，其定价子问题被呈现为通过定制标签校正算法解决的资源受限最短路径问题。我们研究了在中断情况下完全和部分重新规划的利弊，并提供了多式联运网络的管理见解。通过咨询我们的工业合作伙伴为物流网络（包括铁路、水路和公路）生成的真实实例，展示了一组广泛的计算实验。有希望的结果验证了所提出方法的效率，该方法可以轻松调整为实时多式联运物流重新规划。