Bifurcation and linear analysis of connected heterogeneous traffic flow model under cyberattacks

Connected vehicles (CVs) serve as a crucial transitional instrument in the evolution from human-driven vehicles (HDVs) to fully connected autonomous drive vehicles. Their primary role is to enhance driving performance through integration with driver assistance systems (DAS), which utilize vehicle-to-vehicle (V2V) communication for automated computation and analysis to support safer and more efficient driving operations. However, the information interaction between CVs remains vulnerable to cyberattacks (CAs), while driver's subjectivity may further reduce reliance on DAS. To address these challenges, this study introduces a novel compensation mechanism based on the driver's actual solid visual angle and velocity difference. The proposed approach considers both the intensity of CAs and the CV driver's compliance rate to DAS, thereby optimizing the solid visual angle car-following model. Through linear stability analysis and Hopf bifurcation theory, the study derives traffic flow stability conditions and equilibrium points. A bidirectional comparison method is then employed to determine the optimal compensation coefficient, effectively mitigating traffic bifurcation induced by CAs. Numerical simulations, incorporating with parameter calibration based on NGSIM data, were conducted on heterogeneous traffic flows comprising 100 CVs and HDVs. The results demonstrate that appropriate compensator configurations can significantly counteract the adverse effects of CAs without altering equilibrium points. Furthermore, the stability of mixed traffic flows exhibits dependence on CV market penetration rates, with higher penetration rates correlating with stronger resilience against CAs.