The impact of cyberattacks on dynamic behaviors in intelligent mixed car-following models

In intelligent mixed traffic flows with the coexistence of human-driven vehicles (HDVs) and intelligent connected vehicles (ICVs), the communication-enabled features of ICVs render them susceptible to cyberattacks. To investigate the impact of cyberattacks on the dynamic behavior of intelligent mixed traffic flow, this study develops an intelligent mixed traffic flow model under cyberattack scenarios. This model quantifies the impact of cyberattacks targeting the state information of ICVs (including headway and velocity differences within delay times) on traffic flow stability. String stability analysis is employed to examine the impact of attacks on headway of ICVs. Subsequently, to intuitively characterize how different types of attacks and control parameter affect system stability, we innovatively adopt the Definite Integral Stability (DIS) method to calculate unstable roots of the characteristic equation, thereby deriving the instability delay interval of the intelligent mixed traffic system under cyberattacks. Theoretical analysis and numerical simulations demonstrate that cyberattacks severely undermine the stability improvements benefits brought by the penetration of ICVs. Under current model parameters, attacks targeting headway pose a substantially higher risk than those targeting velocity difference. Furthermore, collusion attacks exert far more detrimental effect on traffic flow stability compared to individual attacks. Importantly, incorporating the time-delayed velocity difference effect into the model can effectively mitigate the degradation of system stability caused by cyberattacks.