A review of autonomous intersections: design and management
Abstract
[Background] Autonomous intersections are a new type of urban road intersection in intelligent networked environments. Autonomous intersections replace traffic signals using real-time communication and cooperation of vehicles to realize safe and efficient passage, and are regarded as a vital direction for future urban traffic development. [Objective] To explore the research progress of autonomous intersections, relevant research results are sorted out and summarized in terms of system design and traffic management. [Methods] First, the spatial layout and conflict construction methods of autonomous intersections are introduced. Second, common simplification assumptions made in the research are summarized. Based on the conflict construction methods and assumptions, the system architecture and theoretical model of autonomous intersection management are summarized. Additionally, two commonly used evaluation techniques are presented. Finally, potential directions for future research are proposed. [Results] The spatial layout of an autonomous intersection includes the communication and planning area, control execution area, and conflict area, in which the conflict area can be constructed using the grid method, conflict-point method, or conflict-region method. Autonomous intersection management methods have four modeling approaches: rule-based, optimization-based, learning-based, and game-based. To simplify modeling, studies often make assumptions regarding the vehicle characteristics, communication characteristics, speed and lane management,and environmental conditions. In addition to basic autonomous intersection management for connected and automated vehicles, extended studies include intersection management in mixed traffic environments, pedestrian crossing management, multi-intersection cooperative control, and intersection layout optimization in road networks. Autonomous intersection management has the potential to effectively improve traffic efficiency at intersections, ensuring traffic safety, enhancing driving comfort, and sustainable development of transportation systems. Future research can attempt to further improve the model accuracy, promote the integration of management methods, and enhance the realism of the evaluation environment. [Application] This review provides theoretical support and reference for the design, management, and practical application of autonomous intersections.
