A Max Pressure Approach to Urban Network Signal Control with Queue Estimation using Connected Vehicle Data

Abstract

Max pressure (MP) is a distributed strategy for adaptive urban traffic signal control. Real-time queue estimation for road links is indispensable for MP-based traffic control. All works conducted so far on MP traffic signal control assumed that accurate information of vehicle queues was directly available in real time. This paper studies joint queue estimation and MP control for signalized urban networks with connected vehicles. For the sake of practical significance, the cases of link queue estimation and lane-wise queue estimation were both considered as input to the MP traffic signal control. A congested 3*3 network was emulated using AIMSUN to evaluate the performance of the developed queue estimation and MP traffic signal control algorithms, with study results reported.