Destroying Phantom Jams with Connectivity and Automation: Nonlinear Dynamics and Control of Mixed Traffic

Abstract

Connected automated vehicles (CAVs) have the potential to improve the efficiency of vehicular traffic. In this paper, we discuss how CAVs can positively impact the dynamic behavior of mixed traffic systems on highways through the lens of nonlinear dynamics theory. First, we show that human-driven traffic exhibits a bistability phenomenon, in which the same drivers can both drive smoothly or cause congestion, depending on perturbations like a braking of an individual driver. As such, bistability can lead to unexpected phantom traffic jams, which are undesired. By analyzing the corresponding nonlinear dynamical model, we explain the mechanism of bistability and identify which human driver parameters may cause it. Second, we study mixed traffic that includes both human drivers and CAVs, and we analyze how CAVs affect the nonlinear dynamic behavior. We show that a large-enough penetration of CAVs in the traffic flow can eliminate bistability, and we identify the controller parameters of CAVs that are able to do so. Ultimately, this helps to achieve stable and smooth mobility on highways. History: This paper has been accepted for the Transportation Science Special Issue on ISTTT25 Conference. Funding: This work was supported by the University of Michigan’s Center for Connected and Automated Transportation [U.S. DOT Grant 69A3551747105]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/trsc.2023.0498 .