Development of Non-Recurrent Queue Management System with Connected Vehicles

Road lanes blocked by stopping or slow vehicles or traffic incidents generate non-recurrent queues, which result in bottlenecks and traffic congestion and increasing risk of incidents. Due to huge speed differences between the queue and the adjacent lanes, it is difficult for the queued vehicles to make lane changes to leave the queue. Assisting vehicles in non-recurrent queues to make lane changes earlier is critical to improve mobility of roads. In this paper, a nonrecurrent queue management system is developed with the help of connected vehicles to overcome this challenge. The system utilizes connected vehicles to detect and monitor the status of non-recurrent queues. A game theoretic based lane changing model is applied to estimate optimal speed instructions for connected vehicles to generate larger gaps for queued vehicles to have high probabilities of successful lane changes to the adjacent lanes. The system also provides optimal lane instructions for connected vehicles far behind queues to avoid non-recurrent queues so as to reduce the number of vehicles entering the queues. The system is also evaluated with both synthetic simulations and microscopic traffic simulations, and the analysis indicates that the system can reduce the travel time delay for connected vehicles by up to 84%, and 75% for human-driven vehicles. In addition, the impact of market penetration rates of connected vehicles and the demand levels are analyzed to understand the robustness of the proposed system.