Reservation-based traffic signal control for mixed traffic flow at intersections

Mixed traffic flow involving connected and automated vehicles (CAVs) and human-driven vehicles (HDVs) will be a new state in the future traffic pattern. However, the inherent "stop-and-go" mode of existing traffic signal control seriously weakens the advantage of CAV cooperative operation in mixed traffic. To handle this problem, this study proposes a novel traffic signal control scheme that addresses a mixed traffic scenario at a typical intersection. Specifically, the control scheme can precisely match traffic demand and space-time supply. Then, a protocol of white phases is proposed to decrease the crossing time of CAVs. Finally, the traffic signals and vehicle trajectories are optimized in a unified control framework that minimizes the travel delay of all vehicles. The proposed scheme is verified in SUMO-based simulation considering various penetration rates of the CAVs, different input traffic volumes, and different proportions of leftturning traffic. The experimental results show that the proposed scheme is superior to existing actuated signal control regarding stability when facing different input traffic volumes and proportions of left-turning traffic. Moreover, it is observed that the travel delay decreases with the penetration rates of the CAVs, indicating the potential of the proposed method in the near future mixed traffic scenario.