Critical Traffic Control Locations for Emergency Evacuation

Abstract

Proper traffic control and guidance are imperative to reduce fatalities and property damage. When resources are scarce during emergency evacuation for any natural or artificial disasters, it is crucial to identify the most critical traffic control locations or intersections to deploy temporary control devices or to arrange manual traffic guidance so that the best system performance can be achieved. To answer the urgent theoretical and practical deployment problems, this study has proposed mixed integer nonlinear programming (MINLP) models for both static and dynamic situations. The proposed models aim to find the most crucial intersections, the optimal traffic control strategies for those intersections, and the best destination choices for evacuees to minimize the total system cost. The proposed models assume that evacuees will follow their past travel experience to safe zones unless they encounter the critical intersections. At those fully controlled critical intersections, evacuees will completely comply with the evacuation guidance to help achieve the minimum total system cost. The proposed MINLP models have also been expanded to include cell-based modules to capture the dynamic nature of evacuation flows, which is essential for real-time operation. Numerical tests on a large network with multiple origins and destinations have shown that the proposed models perform reasonably well.