Responding to the Unexpected: Model and Solution Strategy for Combined Dynamic Evacuee Routing and Officer Deployment
Saif Eddin Jabari, Xiaozheng He, Henry X. Liu · Transportation Research Board 88th Annual MeetingTransportation Research Board · 2009
This study proposes a mixed 0-1 integer programming extension to the cell transmission model (CTM) based dynamic system optimal (DSO) problem to identify critical network locations under no-notice evacuation. The formulation developed serves as a baseline analytical model for combined dynamic evacuee routing and improved network throughput through deployment of a limited number of police officers at the critical locations. A heuristic framework is then developed for quick computation, which is of paramount importance under no-notice evacuations. The heuristic scheme uses genetic algorithms to generate officer deployment solutions, while the heuristic algorithm for staged traffic evacuation (HASTE) is used to approximate solution fitness. A numerical example demonstrates the quality of the heuristic framework compared to a solution to the analytical model. Computation times using the heuristic framework are also shown to be significantly smaller than that obtained analytically.