Responding to the Unexpected: Development of a Dynamic Data-Driven Model for Effective Evacuation

Abstract

This research proposes a framework for real-time traffic management under emergency evacuation. A theoretical framework is first proposed for adaptive system control that involves control updating based on real-world traffic data. A heuristic solution framework is then developed to address the computation complexities that come with real-time computations of evacuee routing strategies that aim at minimizing total evacuee exposure time to harm. Further improvements to network traffic throughput are also considered by incorporating officer deployment strategies to critical network intersections. A genetic algorithms based solution scheme is proposed for the combined evacuee routing and officer deployment problem. An evacuation software tool is developed with embedded GIS capabilities that allows users to build evacuation scenarios and run the developed heuristic algorithms. Finally, the quality and efficiency of the developed solution techniques are demonstrated via hypothetical real-world size evacuation scenarios using the software tools.