Abstract: Emergency supplies allocation is an important problem in emergency management. However, uncertain information of the disaster makes it hard to allot materials accurately and efficiently.In this paper, by focusing on the description and update of uncertain information, emergency supplies allocation strategy when earthquake occurs is studied. To describe the disaster, five factors are selected and different situations are constructed based on historical data. The information update process is described as the acknowledgement of the probability of each factor. Based on the updating of disaster information, the rescue process is divided into 2 stages: black stage and grey stage. In black stage, deciders only know magnitude and epicenter of this earthquake, therefore a robust model based on different situations is built. This model generates a robust result that considers the worst situation. Also, in this stage, historical information is used to generate situations to support robust model. In grey stage, damage index of each area updates successively, so that seismic intensity circle can be updated accordingly, and because of which, demand for each kind of material of each area can be updated. Then unsatisfying rate is computed, and a programming model is proposed for following supply allocation. Finally, by comparing our model against a model without information updating, experiment results show that our model has its advantages.

Key words: supplies allocation, two-stage model, robust model, Intensity circle dynamic update