Robust Optimization for Vehicle Routing Problem Under Uncertainty in Disaster Response

The robust optimization problem for emergency logistics in response to disaster is studied in this paper. An improved mixed integer programming model for capacitated vehicle routing problem (CVRP) is developed, which minimizes the total travel cost and unmet demands. Due to the suddenness and unpredictability of disasters, precise information is usually unavailable especially in the early period after the occurrence. Hence the concept of robustness is adopted to handle the uncertainty. The information about demands and travel costs is represented in intervals, which can be estimated based on the historical data. Numerical experiments are conducted and the computational results reveal that the results of the robust optimization model can better balance the total cost and the risk due to the uncertainty.