Zonotope-Based Asynchronous Fault Detection for Markov Jump Systems Subject to Deception Attacks via Dynamic Event-Triggered Communication

By Markov jump systems (MJSs) modeling, this article investigates event-triggered asynchronous fault detection (FD) for quarter-car suspension system (QCSS) via zonotopic residual evaluation (RE). To save communication resources, a stochastic dynamic event-triggered scheme (DETS) is developed to transmit necessary sampled signals while considering deception attacks (DAs) during signal transmission. By means of the hidden Markov model (HMM) with partially accessible mode information, the asynchronization phenomenon between the original system and the FD filter (FDF) as well as the event generator is fully characterized. In such a framework, a mixed
$l_{1}/H_{\infty }$
asynchronous FDF design criterion is derived to guarantee the resultant residual system is sensitive to system fault while being robust to amplitude-bounded exogenous disturbances and measurement noise. Compared with existing results, the proposed approach avoids the mode synchronization limitation and the requirement of ideally known probability information in the transformed probability matrix and conditional probability matrix, and thus can be better applied to practical engineering while covering some previous works as special cases. Moreover, for the first time, a novel zonotope-based dynamic threshold evaluation approach is developed for FD in MJSs instead of the traditional constant threshold scheme for RE. Such an algorithm is designed to effectively avoid inappropriate human empirical threshold selection, which can both shorten the FD time and avoid false alarms hence ensuring satisfactory FD performance. Finally, a quarter-car suspension system is utilized to demonstrate the effectiveness of the proposed design strategy.