Data-driven Agent-based Models for Optimal Evacuation of Large Metropolitan Areas for Improved Disaster Planning

Kazi Ashik Islam (Biocomplexity Institute and Initiative & University of Virginia), Madhav Marathe (Biocomplexity Institute and Initiative & University of Virginia), Henning Mortveit (Biocomplexity Institute and Initiative & University of Virginia), Samarth Swarup (Biocomplexity Institute and Initiative & University of Virginia), Anil Vullikanti (Biocomplexity Institute and Initiative & University of Virginia)

Abstract

Evacuation plans are designed to move people to safety in case of a disaster. It mainly consists of two components: routing and scheduling. Joint optimization of these two components with the goal of minimizing total evacuation time is a computationally hard problem, specifically when the problem instance is large. Moreover, often in disaster situations, there is uncertainty regarding the passability of roads throughout the evacuation time period. In this paper, we present a way to model the time-varying risk associated with roads in disaster situations. We also design a heuristic method based on the well known Large Neighborhood Search framework to perform the joint optimization task. We use real-world road network and population data from Harris County in Houston, Texas and apply our heuristic to find evacuation routes and schedules for the area. We show that the proposed method is able to find good solutions within a reasonable amount of time. We also perform agent-based simulations of the evacuation using these solutions to evaluate their quality and efficacy.