Single-Lane Roundabout Manager under Fully Automated Vehicle Environment

Abstract

In the advent of new automation and communication technologies, the introduction of connected and autonomous vehicles (CAVs) can make efficient use of the space on our road networks. Since roundabouts suffer from lower efficiencies when traffic flows are unbalanced and unfamiliar drivers are present, CAVs have the potential to improve traffic operations, reducing lost time and inefficiency in gap-usage. The aim of this research is to take advantage of those recently available technologies to design a traffic management system for roundabouts. Such a system can provide CAVs with optimal trajectories to negotiate the roundabout, maximizing throughput and minimizing control delay. The rule-based Roundabout Manager algorithm developed considers the roundabout as a whole system, assessing the priority of incoming vehicles on a first-come-first-served scheduling to solve conflicts. The management system takes into consideration the geometry of the roundabout, including the flared entries and exits and the location of conflict sections based on merging trajectories of vehicles in negotiating the roundabout. To illustrate the operation of the algorithm, a total of 15 undersaturated demand scenarios with 10 replications per scenario were simulated. From the results, it can be concluded that the system can reduce control delay for medium demand scenarios, maximizing the throughput for all demand scenarios. The Roundabout Manager can work in real time providing optimized trajectories in less than 0.02 s at any traffic flow rate simulated.