A two-stage chance constrained stochastic programming model for emergency supply distribution considering dynamic uncertainty

This paper presents a comprehensive approach to addressing the challenges of designing a reliable emergency logistics network under the dynamic uncertainty of natural disasters. The proposed multimodal transportation strategy based on level assignments can effectively mitigate the risk of facility disruptions and network uncertainty and enhance the adaptability of the system to diverse emergency supplies. The established two-stage chance constrained stochastic programming model and improved evolutionary algorithm are powerful tools for optimizing the network design and improving its resilience. Based on the real-world case of the Ya'an earthquake in China, the impacts of various practical constraints are analyzed. Several insights into the decision-making process of emergency logistics management are provided: managers should prioritize time utility over cost utility and use multiple aircraft and helicopters, supplies should be sorted into fewer and more important categories and placed in smaller packages for efficient transportation, multimodal transportation should be considered based on geographical and traffic conditions.