Abstract: There would always be a lot of crisis locations when an unconventional emergency breaks out. The requirements of each crisis location are usually different, which is difficult to meet the requirements of multiple crisis locations for a single resource centre. So it is a practical problem to be solved urgently by decision makers how to fairly and reasonably schedule emergency resources for multiple crisis locations. According to the demand information, the dynamic process of emergency resources scheduling for multiple crisis locations are described, in which the emergency resources scheduling process are divided into several stages according to the change of demand information for multiple crisis locations. On this basis, a theoretical model of multi-stage emergency resources scheduling process is designed for multiple crisis locations. After a series of assumptions are made, the game model based on resource requirements of multiple crisis locations is set up by using game theory according to the degree of disaster, and the improved ant colony optimization (ACO) is introduced to seek out the solution in order to schedule emergency resources for multiple crisis locations according to the minimum virtual cost. Simulation tests and numerical analyses are given to demonstrate the feasibility and availability of the model. The model and algorithm can also provide a new solution and approach for the distribution of resources in business logistics.

Key words: unconventional emergency, resource requirements, game scheduling, improved ACO