Abstract
Routing problems have been extensively studied in operations research because of their wide application in transportation and logistics. In this study, we consider routing problem in a road network of which the traveling conditions change over time and sometimes are uncertain. Such problems can arise in humanitarian logistics, resident evacuation, and emergency resource delivery after natural disasters and severe weather events. We provide a methodology to support route planning with limited information of the network conditions and route updating as new information becomes available. The dynamic network condition is modeled by a time-varying speed reduction factor. We update the estimation of this factor by integrating prior estimation with the latest travel data from the vehicles in a Bayesian inference framework. Ant colony optimization method is used to find the optimal route in both planning and updating phases. A real-world case study shows the effectiveness and necessity of considering dynamic uncertain network conditions in routing.
Keywords: Dynamic decision making; Bayesian inference; Vehicle routing; Ant colony optimization