Improved reservation-based method with controllable gap strategy for vehicle coordination at non-signalized intersections

Abstract

Recently, new research has emerged for signal-free intersection control via V2X technologies, and reservation-based intersection control for connected and automated vehicles (CAVs) has great potential to improve travel efficiency and traffic safety. A centralized traffic controller is the key component to coordinate movements of the approaching vehicles considering their priorities. In this paper, we propose a controllable gap strategy for the traffic controller considering the conflict relationship. To reduce the number of timestamp comparisons, we build three conflict modes of different traffic movements and present safe speed-dependent constraints for them. Then, based on the interactive relationship, we design an algorithm to manage the reservations for vehicles, which allows conflict-free vehicles crossing over the intersection synchronously. Furthermore, to ensure absolute safety in all cases, we present a rapid check criterion for conflict-free time slots in the conflict set of the reservation to avoid a potential collision, which removes unnecessary judgments. It is analyzed and proved that the aggregation of these local checks solves the conflicts globally. The strategy relaxes the limits of reservation policy. The proposed method is verified in the SUMO platform compared with the naive reservation-based method. Simulation results within 3000 s show that the proposed method can increase the traffic throughput and avoid vehicle collision for isolated intersections and traffic networks.