Traffic Control Strategy Formulation and Optimization Enabled by Homogenous Connected and Autonomous Vehicle Systems.

Abstract

Traffic congestion has become a serious issue all over the world due to the rapid increase in population and traffic demands. The advances in Connected and Autonomous Vehicles (CAVs) have demonstrated a potential to improve traffic mobility and safety performance at intersections. An advanced intersection control system, CAV-enabled Intersection Management Mechanism (CAVIMM), was developed to ensure traffic safety and operation efficiency at intersections without using any traditional traffic signals. CAVIMM releases vehicle movement restrictions of dedicated turning lanes in traditional signalized intersections by enabling left-turn, straight and right-turn movements from any lane of each approach. CAVs approaching intersections are controlled by an Intersection Control Center (ICC) using communication between vehicles and intersection infrastructures (V2I). Due to the increasing number of potential conflict areas introduced by CAVIMM, there is a need for a precise vehicle trajectory model, which helps the ICC to better control CAVs passing through the intersection. A trajectory coordination model, Temporal-Spatial Dimension Extension-based Trajectory Coordination Model (TSDTCM), was developed to account for CAV widths and lengths. Based on vehicle trajectory formulation, various algorithms for determining conflicting CAVs passing sequences were developed. The First-Come-First-Serve (FCFS) algorithm is the most widely used one due to its simplicity; however, it may not be the optimal one in terms of minimizing total intersection delay or maximizing overall throughput. Therefore, two CAVIMM systems using FCFS and an optimal algorithm, Discrete Forward-Rolling Optimal Control (DFROC) algorithm, were developed in this study. The performance comparisons between CAVIMM systems and traditional signal control were conducted at a 4-leg intersection through VISSIM-based simulations. Their performance under v different scenarios with various traffic volumes and intersection configurations were evaluated. Experimental results indicated that the CAVIMM system outperformed traditional traffic signals in terms of reducing total traffic delay at intersections. At the 100% level of volume, the total traffic delay was reduced by more than 90% for all scenarios of CAVIMM system. Several factors, including intersection configurations, traffic volumes, algorithms of determining CAVs’ passing sequences and the sizes of buffer zone, directly affected the CAVIMM performance with respect to the reduction of the total traffic delay.