TLB-VTL: 3-Level Buffer Based Virtual Traffic Light Scheme for Intelligent Collaborative Intersections

To improve the safety, traffic efficiency, and fairness among vehicles at intersections, it is urgent to study intelligent collaborative strategies and make intersections smarter. In this paper, a 3-Level Buffer (TLB) based Virtual Traffic Light (VTL) scheme, named TLB-VTL, is proposed for intelligent collaborative intersections. The intersection is divided into three adaptive areas according to the traffic flow of each lane, and the sequence of each timing cycle is calculated in realtime according to the flow in TLB around an intersection. To ensure fairness, the difference in probability of each lane to pass an intersection is restricted to a lower level. The VTL is realized based on communications of vehicle-to-vehicle (V2V), vehicle-to-roadside (V2R), and vehicle-to- infrastructure (V2I), which could improve the safety and fairness without involving traffic lights. Moreover, a Cooperative Collision Avoidance Predictive control (CCAP) algorithm is proposed, which can assist vehicles to go across the next intersection without stopping through predicting the time conflict and generating an efficient traffic schedule for the entire road network. The simulation results indicate that the proposed TLB-VTL algorithm improves the fairness by 331%, decreases the average delay by 88%, and improves the ability to solve congestion by 12% compared with the traditional traffic light algorithm. Besides, the CCAP algorithm increases the traffic fluency by 45% at the intersection.