Optimal Traffic Routing for Large-Scale Evacuation in Urban Networks with Various Threat Levels

This paper investigates a traffic network evacuation model with non-uniform threat levels, minimizing an objective of total exposure (duration and severity) to the threat among all vehicles. The model is essentially a min-cost dynamic flow problem associated with time-dependent arc costs. The traffic flow models included in this representation are the point queue and the spatial queue models. The evacuation model is solved to optimality on a time-expanded network using commercial software. The model and associated solution method are applied to solve a large-scale chlorine release scenario in Tucson, Arizona.