Simultaneous Optimization of Traffic Signal Control and Vehicle Platooning Scheme Based on Connected and Automated Vehicle (CAV) Technology

Connected and automated vehicle (CAV) technologies can improve the traffic operation and management in urban transportation systems. In particular, a great advantage is expected when CAV technologies are applied to the traffic signal control for intersections. So far in the academy, researchers have dealt with platooning vehicles moving in an approach to an intersection to reduce stops and delays without traffic signals. The present study focuses on synchronizing stopped vehicles at a stop line in severely congested traffic condition. If vehicles in a queue move synchronously when green light initiates, the traffic throughput can be maximized without start-up delays. A reinforcement learning (RL) algorithm is devised to simultaneously optimize both the traffic signal control and the platooning scheme in a unified modelling framework. The proposed approach outperforms both an adaptive traffic signal control without platooning and a fully automated intersection management (AIM) without traffic lights, as traffic flows are oversaturated.