Modeling and analysis of traffic flow with automated vehicles affected by information deviations

Information deviations are inevitable under the influence of multifarious factors in real-world traffic, leading to discrepancies between the information obtained by automated vehicles (AVs) and the true information. However, due to the lack of an appropriate analytical model that incorporates various information with deviations, we have limited knowledge of the relationships between different types of information deviations and anomalous dynamics of AVs and traffic flow. This study aims to fill this gap. Specifically, we first expound the possible information deviations in AVs, upon which we categorize them into three types: velocity, gap, and driving decision deviations. Subsequently, we modify the input parameters in the adaptive cruise control (ACC) model that calibrated using real experimental data to capture the car-following dynamics of AVs with information deviations. By using \(H_\infty\) $H_{\infty }$ control theory and characteristic equation-based method, we derive the local and string stability criteria of traffic flow with AVs, so as to discern the effects of various system factors on traffic flow stability. Experimental results show that information deviations could provoke abrupt acceleration or deceleration of AVs, leading to instability in automated traffic flow, oscillation, and even collision accidents. Overall, this paper unveils the influence mechanisms of diverse information deviations on AVs and traffic flow, providing valuable suggestions and theoretical guidance for the future development of AVs.