A computationally efficient and refined signal control method for isolated intersections in a connected vehicle environment

The advent of connected vehicles (CVs) has enabled the availability of richer and more accurate data for more flexible and sophisticated traffic signal control. However, the complexity of optimization models, especially when individual vehicle trajectories are considered, makes solving existing CV-based signal control methods difficult. This study proposes a computationally efficient and refined signal control method for isolated intersections in a CV environment. A mixed-integer nonlinear program model, which minimizes the total vehicle travel time, is developed by employing a simplified car-following model to predict individual vehicle trajectories. To address the computational concerns, the signal optimization model is reformulated by clustering incoming vehicles into platoons and analyzing the platoon features based on the interactions between vehicles as well as between vehicles and traffic signals. Simulation studies are conducted to compare the proposed method with the adaptive signal and vehicle-actuated control methods. The results show significant reductions in vehicle travel time and fuel consumption compared to benchmark methods under different demand levels. Furthermore, the proposed control method has high computational efficiency with a solving time of less than 0.05 s, validating its potential for practical applications.