Critical Intersection Signal Optimization During Urban Evacuation Utilizing Dynamic Programming

Abstract

Despite advancements in the field of traffic planning and operations, many major cities still rely on pretimed signal settings. With only pretimed signal control strategies, the secondary effects of a terror attack are magnified by slow evacuation times resulting in further loss of life. However, the cost of implementing new infrastructure based solely on the chance of a no-notice evacuation is not something that city planners are willing to do. The purpose of the research is to define a cost-effective methodology to develop pretimed signal control strategies to assist evacuations in urban areas. To that end, a dynamic programming methodology was developed to assist critical intersections in urban corridors. To test this methodology, a microscopic traffic-simulation environment was created for a case study of a 10 intersection evacuation corridor in Washington, DC. Using the proposed methodology to optimizing signal splits of a critical intersection within an evacuation corridor, evacuation clearance time was reduced by approximately 1 h. Furthermore, this formulation can be used to develop pretimed signal control settings for evacuation scenarios. Results showed that peak-hour signal timings are not sufficient in the case of an emergency, and signal timing plans must be tailored for emergency evacuation.