Optimal Access Management for Cooperative Intersection Control

This paper presents an intersection access management methodology that optimizes the crossing sequence of an automated intersection, where the low-level vehicle control is performed by a cooperative intersection control (CIC) strategy. While the CIC regulates the safe and efficient relative motion of vehicles in the intersection, a high-level hybrid queuing model is proposed to describe the dynamics of the vehicle queues associated to each intersection lane. This model, including constraints, is used to design an optimal access management approach based on the model predictive control that minimizes the time that the vehicles spend within the intersection, thereby optimizing the traffic throughput of the intersection. The performance of this methodology is studied by means of two representative examples. The impact of the design parameters of the optimal access management approach is shown for a T-intersection case study. Moreover, using a real-life five lane intersection case study, the proposed approach is compared to a vehicle-actuated traffic light approach, and a first come first served approach. The comparison shows the benefits of the automated optimal serving of vehicles from different lanes.