Abstract: In real-world emergency responses, helicopters are gradually used to transport key relief resources such as medical supplies and personnel, due to the long distance between supply nodes and demand nodes or the difficulty to reach to affected areas by land in time. However, it is often impossible to transport relief supplies to every medical aid point (MAP) by helicopters in large-scale disasters. In general, considering the aggregation of residents in disasters, emergency distribution points (EDPs) are set up for receiving urgent medical supplies from helicopters and then delivering the supplies to each MAP by vehicles. In the work, a clustering-based two-stage approach is developed for medical supplies intermodal transportation in disasters. In the first stage, according to the locations of MAPs, Fuzzy C-Means (FCM) is used to select EDPs and partition MAPs, and importantly an improved FCM algorithm with capacity constraints (FCMwCC) is developed to overcome the unbalanced extra capacity in the FCM partitioning, which can determine the network of helicopters-vehicles intermodal transportation. In the second stage, a clustering-based routes optimization model is developed, which can determine the vehicle traveling routes from EDPs to their covered MAPs. Numerical experiments test the effectiveness of the developed approach and algorithms, and the following findings are obtained. Firstly, the MAPs partitioning by the classic FCM could minimize the total distance among MAPs and EDPs, but could result in big extra capacities due to lack of considering the capacity constraints. Further the MAPs adjustment by the improved FCMwCC could decrease the extra capacities and the number of used medical vehicles. The proposed model and algorithms could select proper emergency distribution points and arrange transportation routes for real-world intermodal transportation of medical supplies.

Key words: large-scale disasters, medical supplies, intermodal transportation, temporary distribution centers selection, routes optimization