A Molecular Dynamics-based Car-following Model for Connected and Automated Vehicles Considering Impact of Multiple Vehicles

The car-following control of connected and automated vehicles (CAV) is studied under mixed traffic flow with regular vehicles. Considering the factors of velocity, headway, and the velocity and acceleration difference of multiple front and rear vehicles, this paper constructed a car-following model for CAVs by using the molecular dynamics to express the impacts of different surrounding vehicles on the host one. The car-following process of CAVs was then described when driving in the traffic mixed with CAVs and regular human-driving vehicles. The results of stability analysis show that, compared with the full speed difference model, the proposed model is more conducive to improving the stability of traffic flow due to the consideration of multiple front and rear vehicles' information. The simulation results indicate that, compared with the Cooperative adaptive cruise control (CACC) model provided by PATH laboratory, the average maximum error of our model is reduced by 0.19 m/s, the average error is reduced by 26.79%, and the fitting accuracy is improved by 0.91%. Besides, in the traffic flow mixed with CAV and RV, with the increase of CAV penetration, the average velocity of the platoon and volume increase gradually. The results of Hysteresis loops show that compared with the Full Velocity Difference (FVD) model, the stability of traffic flow under the CAV model in this paper is better. The proposed model can serve as an effective method for CAV control under both homogeneous traffic flow or heterogeneous flow mixed with CAVs and human-driving vehicles. Under the situation that it is difficult to carry out the field test of heterogeneous flow mixed with CAVs and human-driving vehicles, this study provides a theoretical basis and model support for vehicle control as well as infrastructure planning and design. Copyright © 2022 by Science Press.