Multi-objective Emergency Scheduling for Geological Disasters

Because of the specific time and distance constraints, emergency management departments usually build multiple depots (resource centers) to serve the widely dispersed customers (disaster areas), to more effectively fulfill customer demand, and deal with the changing road conditions in real time. Therefore, research on multi-depot dynamic emergency dispatch can be of significant value to effective disaster operations. In this paper, a multi-objective multi-depot and multi-type dynamic vehicle routing problem model is established that minimizes total distance and priority errors and considers secondary disasters, damaged roads, the limited vehicle number at the various depots, and the different vehicle capacities. To solve this model, a hybrid ant colony optimization based on the circumcenter of the polygon formed by depots is proposed. Real landslides disaster data from Hubei province and two kinds of benchmark instances test the performance of the proposed algorithm. After detailed experimental comparisons, the competitive performance of the proposed algorithm is verified.