A car-following model to assess the impact of V2V messages on traffic dynamics

Abstract

Connected vehicles (CVs) are considered to have the potential to significantly improve traffic flow stability. Although several studies have been devoted to modelling car-following behaviour in a connected environment, most model formulations are based on assumptions without empirical observations. Therefore, this paper utilizes data from field experiments to explore the dynamics of CVs. Data mining analysis shows that the driver is more responsive to velocity differences with safety messages. According to the data analysis results, we present a modified car-following model based on the intelligent driver model (IDM). Then, the parameters of our modified IDM are calibrated. It is shown that the modified IDM is able to reproduce the observed experimental data better than the original IDM. Next, we conduct a linear stability analysis of the modified IDM to explore the properties of the model. Finally, simulation experiments are conducted to verify the theoretical analysis.