Phase transition in lattice hydrodynamic model accounting for information interruption under connected autonomous vehicles conditions

As the Internet of Vehicles (IoV) becomes deeply integrated with autonomous driving technologies, connected and autonomous vehicles (CAVs) leverage V2X communication to achieve multi-source information exchange and multi-vehicle cooperative control, demonstrating significant advantages in mitigating traffic oscillations and enhancing traffic capacity. However, information interruptions caused by signal interference and network congestion in the actual IoV environment alter the vehicle information acquisition and feedback mechanisms, weakening traffic flow stability and inducing congestion phase transitions. Existing research lacks a systematic characterization of the mechanisms underlying traffic instability in CAVs under non-ideal communication conditions. To address this gap, we in this paper develop a lattice hydrodynamic model that incorporates predictive effect and flow difference effect under information interruption situation, which distinguishes between vehicle information acquisition and feedback modes in normal communication and interrupted state. Through linear stability analysis, nonlinear analysis and numerical simulations, the study systematically investigates the mechanisms how information interruption effect affects the stability and phase transition behaviors of CAVs. The results indicate that an increase in the information interruption probability significantly reduces traffic stability and constricts the linear stability region. Under normal communication conditions, multi-vehicle predictive feedback and flow difference feedback can effectively enhance system stability. In contrast, during communication interruptions, single-vehicle predictive compensation and flow difference compensation can partially suppress instability, albeit with considerably weaker effects than their multi-vehicle cooperative counterparts under normal conditions. These findings suggest that high-reliability communication and wide-area cooperative perception are critical to maintaining stable operation of CAVs. Furthermore, we introduce power spectrum and spectral entropy into the analysis of traffic flow, supplementing the characterization of the impact of information unreliability on system disorder from the frequency domain perspective. The results reveal that the increase of information interruption effect leads to a broader power spectrum and a marked rise in spectral entropy, reflecting heightened the system disorder and diminished the system stability. Conversely, enhancing effective feedback helps to suppress spectral broadening and entropy increase. These frequency-domain findings are consistent with linear stability theory and numerical simulations, thereby deepening the dynamical understanding of phase transitions and congestion evolution of traffic flow induced by information interruption effect. Collectively, the results provide theoretical support for the communication optimization and cooperative control design in intelligent connected transportation systems.