Traffic-adaptive signal control and vehicle routing using a decentralized back-pressure method

Abstract

The problem of controlling traffic lights under adaptive routing of vehicles in urban road networks is considered. Multi-commodity back-pressure algorithms, originally developed for routing and scheduling in communication networks, are applied to road networks to control traffic lights and adaptively reroute vehicles. The performance of the algorithms is analyzed using a microscopic traffic simulator. The results demonstrate that the proposed signal control and adaptive routing algorithms can provide significant improvement over a fixed schedule and a single-commodity back-pressure signal controller, in terms of various performance metrics, including queue-length, trips completed, and travel times.