Project Summary
The objective of this research is to evaluate the impacts of connected and autonomous vehicles (CAVs) on the performance of signalized networks at the aggregate level. CAVs are expected to improve alleviate traffic congestion, but their impacts are usually evaluated at the microscopic level, for example, through the design of optimal vehicle trajectories or optimal operation of individual intersections. In this proposal, we aim to develop a new performance evaluation framework through network fundamental diagrams (NFD), which capture the relationship between the average flow-rate and density at the network level. In particular, we will study how individual advisory speed limits of connected vehicles and different start-up and clearance behaviors of autonomous vehicles can increase the network capacity and reduce the start-up and clearance lost times. This research will take advantage of a microscopic simulation platform based on Newell's car-following model that incorporates different bounded acceleration (start-up) and aggressiveness (clearance). Connected vehicles' individual advisory speed limits are determined by a feedback control strategy which incorporates traffic signal information and loop detector data. Different start-up and clearance behaviors of autonomous vehicles are implemented by changing their acceleration bounds and aggressiveness when traffic lights turn yellow. Under each combination of technologies, we determine the NFD in time-independent stationary states with periodic vehicle trajectories. A goal is to determine whether their impacts are additive or alternative.
