Eco-driving trajectory optimization model at signalized intersection considering shared phase

A dynamic programming-intersection conflict management strategy was proposed to optimize the trajectories of intelligent connected vehicles approaching signalized intersections under shared phase conditions and mitigate conflicts at intersections. A dynamic programming model was established based on the information of vehicle state and signal phase and timing to optimize the trajectories of vehicles upstream of the signalized intersection, maximize throughput during green time, and reduce waiting time. Besides, an intersection conflict management strategy was designed for the shared phase scenario with traffic conflicts. The strategy determined the sequence of conflicting vehicles passing through the intersection by virtual vehicle mapping and created a safe gap by an intelligent driver model, ensuring smooth traffic flow within the intersection. Finally, a simulation analysis was conducted on the signalized intersection of Yongqing Road and Yonglong Road in Xi'an City. Simulation results show that in contrast to left-turn protected phase and shared phase scenarios under the control of the dynamic programming model, the proposed model improves average speed by 12.88% and 4.14% and reduces energy consumption per 100 km by 9.79% and 3.97%, respectively. Compared to the scenario with 0% penetration rate, the total energy consumption per 100 km under the proposed model decreases by 3.56%-13.97% in scenarios with penetration rates ranging from 20% to 100%. Analysis of the time to collision and post-encroachment time under the proposed model shows a significant improvement in safety. Furthermore, under conditions of varying traffic demands and fluctuating signal cycles, the proposed model can achieve trajectory optimization for vehicles throughout the entire process from entering the lane to leaving the intersection.