University of California Transportation Center

The events of hurricane seasons and the threat of terrorist attacks have made evacuation during disruptions a leading management issue. Transportation networks, which form the backbone of any emergency management plan, should be able to respond to disruptions by ensuring safe, organized and quick movement of people at the time of crisis. This work proposes two models to capture the highly uncertain and time-dependent nature of transportation systems in the face of an emergency. The first model addresses the location of safety shelters. It uses risk management tools, the classical facility location model and traffic assignment techniques with Wardrop’s principle to determine the optimal location of shelters. The second model deals with real time decision-making during evacuations. It incorporates a simulation algorithm with the successive shortest path algorithm to model evacuation. Detailed traffic information in the network can be obtained from this algorithm to facilitate the evacuation.