In this research, we focus on designing an optimal control framework for guiding a platoon of connected vehicles through a signalized intersection and eva‎luating its performance in mixed traffics (including both connected and human-driven vehicles). Since achieving fully connected traffic in the near future is not feasible, this research could provide an effective assistance for improving traffic in large cities. The proposed optimal framework is divided into two optimization problems: discrete and continuous. The first problem uses the job-shop scheduling method, and the second problem implements the optimal energy or time path for executing the schedule using Hamiltonian analysis. Then, the dynamic controllers for the connected vehicle are designed and eva‎luated using three conventional methods, and one of them is selected for next simulations. In previous research, it was shown that creating platoons of vehicles can help in better planning and reducing planning and movement costs. Therefore, platoons are formed for connected vehicles, with the condition that they share the same route and no human-driven vehicles are present between them. The software SUMO is used for simulation, considering various traffic volumes, rates of human-driven vehicles in traffic, allowable number of vehicles in a platoon, and number of lanes leading to the intersection. This result in multiple simulations, where the total fuel consumption of all vehicles and the exit time of the last vehicle from the intersection are considered as outputs. The results indicate that the proposed optimal framework can reduce both time and energy consumption compared to uncontrolled traffic. However, the effectiveness of the optimal framework diminishes with an increase in the presence of human-driven vehicles in traffic. Additionally, examining the impact of connected vehicle platooning show no clear relationship between platoon size and reduction of travel time or energy consumption. Further studies using data mining methods might reveal a better relation between it and traffic patterns and enhance the performance of the optimal framework.

محمد دانش

حسين فلسفين

مهدي كشميري , فريد شيخ الاسلام