Location and Distribution Decision of Emergency Logistics Network under Uncertain Demand Distributions

Abstract

To ensure the supply of materials after urgent events, including recent public health crisis, it is crucial to formulate a reasonable preventive strategy for the location and distribution of emergency materials within the logistics network. In this paper, a class of multi-category emergency material storage distribution models under scenarios with uncertain post-event demand distribution is studied. Based on historical small-sample post-event demand information, the model constructs a two-stage material storage and scheduling distributionally robust optimization model. By applying a dual method, the non-linear distributionally robust optimization is transformed into linear optimization, and an improved Lagrangian L-shaped algorithm is designed to solve the two-stage model. Finally, the robustness of the model and algorithm is verified by implementing the method to a numerical example, and the sensitivity of the location and  distribution decisions to material shortage under different levels of urgent events is analyzed.