A Novel Prescribed-Time Performance Security Control for Vehicular Platoon Under Dual Attacks via NN-Based Extended State Observers

This article investigates the security control problem for the heterogeneous autonomous vehicle platoon subjected to dual false data injection attacks, namely position sensor attack and actuator attack, as well as external disturbances. To begin with, an exponential spacing policy is introduced to mitigate the time-lag effects of existing spacing policies and further enhance traffic flow efficiency. Then, dual attacks and matched/unmatched external disturbances are modeled as the lumped disturbance for each channel of the system, and a set of neural-network-based extended state observers is developed to accurately estimate these lumped disturbances. Furthermore, to overcome initial error boundary constraints and improve platoon transient/steady-state performance, a modified prescribed-time performance security control method is proposed via a novel switch-driven error transformation function. Based on the methods mentioned above and relying on the backstepping technique, this study ensures the stable and safe driving of the heterogeneous autonomous vehicle platoon. Ultimately, the effectiveness and superiority of the presented approach are validated by simulation cases and comparative analysis results.