Research on Platoon Dynamic Dispatching at Unsignalized Intersections in Intelligent and Connected Transportation Systems

In Intelligent and Connected Transportation Systems (ICTS), vehicles, infrastructure, pedestrians exchange information through two-way telecommunication network. The system collects and uses multi-source data sufficiently so that control measures can be implemented in real-time. However, few studies have discussed how to achieve fairness based on dynamic priority. In this paper, the Platoon Dynamic Dispatching Model (PDDM), based on Vehicle-Intersection Coordination Scheme (VICS), is proposed to control vehicles. The first stage establishes the trajectory model and defines the conflict zones at the intersection. Platoons forms at the upstream, then the conflict risk indicator model is applied to predict the collision risk of the platoons. The dynamic priority dispatching model is applied to change the passing order of platoons at an intersection from all directions in real-time according to their status at the upstream. Finally, PDDM is solved by the wolf pack algorithm. The scenario is simulated by SUMO, and the Python-TraCI Library is used for interfacing a python script. The experimental results show that when the arrival rate is 1200pcu/h, the traffic efficiency is 2.83 times that of the Priority Control Model (PCM) model, and there is almost no parking. The average travel time and the average lost time is significantly reduced. Under the initial priority conditions, the results meet the expectations of the dynamic priority dispatching model.