Evacuation of a highly congested urban city

Abstract

For a given transportation network, evacuation route planning identifies routes and route schedules to minimize the time to evacuate the vulnerable population due to an imminent disaster. In this paper, we present a capacity-constrained routing optimization approach to maximize the total evacuees for short notice evacuation planning in case of both natural and/or man-made disasters. The methodology is scaled to a highly congested urban city with a transportation network capacity well below daily peak demands. As the evacuation route planning is computationally challenging, an evacuation scheduling algorithm was adopted to expedite the solution process. The algorithm uses Dijkstras algorithm to find the shortest path(s) and a modified greedy algorithm to assign maximum flows to selected paths given a specific schedule per time interval. A case study using real population and transportation network data was tested using the proposed methodology. The results show that the city is in grave danger due to the high population density and disproportionate network capacity.