Optimal control method of heterogeneous vehicle traffic flow under Connected Vehicle environment

With the introduction of intelligent connected vehicles (CVs), the mixed traffic of heterogeneous motor vehicles (including human-driven vehicles (HVs) without networking capabilities and intelligent CVs with networking capabilities) may have negative impacts on the transportation system, and the control and guidance of traffic flow for different types of motor vehicles become more complex. Considering the behavioral differences and information acquisition differences among individual vehicles of different types, an optimal control method for heterogeneous motor vehicle traffic flow based on the user equilibrium (UE) model and Bayesian optimization (BO) algorithm is proposed. Firstly, initial traffic volume is assigned to the network based on the UE model, then a Bayesian optimization model is built using the Gaussian process probability surrogate model and acquisition function to optimize the control strategy of CVs for the improvement of heterogeneous traffic flow. The effectiveness of the proposed method is verified using the classic Nguyen-Dupuis network, which demonstrates its good solving efficiency and significant optimization effect.