A Queue-length Threshold-based Control Strategy and Queueing-analytic Model for CAVs at Signal-Free Intersections

The primary challenge in managing signal-free intersections in a connected environment is the allocation of right of way for connected and autonomous vehicles (CAVs) in conflicting directions. The stochastic and unpredictable arrivals of CAVs generally result in uncertainties regarding their queueing process and delays at signal-free intersections, which must be considered in the management of traffic flow. In this paper, we propose a queue-length threshold-based control strategy for the management of CAVs at signal-free intersections, thereby improving traffic effectiveness and reducing delays. This strategy is designed to dynamically adjust the right-of-way based on the fluctuations in CAV queue length at intersections. To explicitly evaluate the performance of the proposed control strategy, we further develop a queueing model and derive some basic traffic performance metrics. In addition, a series of numerical experiments was conducted to illustrate and analyze the effectiveness of the proposed strategy under various traffic conditions. Our results indicate that queueing performance at signal-free intersections can be affected by road storage capacity, traffic volume, and strategy settings.