A Hierarchical Framework for Model-Free Adaptive Control of Heterogeneous Multiple High-Speed Trains With Deception Attacks

This paper studies a hierarchical distributed model-free adaptive control issue for heterogeneous multiple high-speed trains with random deception attacks. By employing the dynamic linearization method, the high-speed train systems are modified to an equivalent linear data model. Then, a novel hierarchical distributed model-free adaptive security control framework is designed mainly including the following perspectives: 1) an adaptive distributed observer is devised for each train to estimate the virtual reference signal; 2) by employing the local information, a decentralized control scheme is developed to ensure that every train can track the reference velocity and position trajectories; 3) a deception attacks detection and compensation strategy is designed to determine whether the received data is attacked or not and compensate for the false signals. The system stability is rigorously proven via the contraction mapping technique and mathematical induction method where the speed and position tracking of heterogeneous trains is guaranteed. Eventually, a simulation example is provided to verify the effectiveness of the proposed protocol.