Traffic Signal Timing Optimization with Connected Vehicles

Thesis (Ph.D.)--University of Washington, 2019The advent and deployment of Connected vehicle (CV) and Vehicle-to-everything (V2X) communications offer the potential to significantly improve the efficiency of traffic signal control systems. The knowledge of vehicle trajectories in the network allows for optimal signal setting and significant improvements in network performance compared to existing traffic signal control systems. This research aims to develop a framework, including modeling techniques, algorithms, and testing strategies, for urban traffic signal optimization with CVs. The objective is to improve the safety, mobility, and sustainability of all vehicles in the study areas utilizing CV data, i.e., real time information on vehicles’ locations and speeds, as well as communications to the signal control systems. The proposed framework is able to optimize traffic signal timing for a single intersection and along a corridor under various market penetration of CVs. Under full penetration rate of CVs, the signal timing optimization and coordination problems are first formulated in a centralized scheme as a mixed-integer nonlinear programing (MINLP). Due to the complexity of the model, the problem is decomposed into two levels: an intersection level to optimize phase durations using dynamic programing (DP) and a corridor level to optimize the offsets of all intersections. Under medium-to-high penetration rates of CVs, Kalman filter methods are applied to estimate trajectories of unequipped vehicles given the available trajectories of CVs. The estimated trajectories combined with CV trajectories are utilized in the trajectory-based signal timing optimization process. Under relatively low penetration rates of CVs, a Deep Intersection Spatial Temporal Network (DISTN) is developed to predict short-term movement-based traffic volumes. The predicted volumes are used in a volume-based adaptive signal control method to calculate signal timing parameters. Comprehensive testing and validation of the proposed methods are conducted in traffic simulation and with real world CV (probe vehicle) data. The testing tasks aim to validate that the developed methods are computationally manageable and have the potential to be implemented in CV-based traffic signal applications in the real world