Dynamic event-triggered asynchronous fault detection for Markov jump systems with partially accessible hidden information and subject to aperiodic DoS attacks

Highlights：

• Based on HMM with PAHI and switched system analysis techniques, a unified framework is established by fully considering the effects of DETS, aperiodic DoS jamming attacks, partially unknown TRs and CPs, and sufficient conditions are developed to guarantee that the resulting switched residual model is stochastically stable and has a weighted dissipativity performance. In such a way, the resource awareness, resilience performance, and dissipativity performance can be analyzed simultaneously. Besides, a quantitative relation among the system parameters, sampling period, DoS duration, sleeping period, and dissipativity performance index is explicitly described in the stability criterion.

• A co-design criterion of asynchronous FDF and resilient-DETS for MJSs with PAHI is deduced, which allows the unknown elements to be existing in the TRM of Markov chain, the CPM of observed signal, or in both of them, and even in the extreme scenario where the TPs and CPs are completely unknown, the proposed method can still provide feasible solutions while existing methods cannot [7–11]. Compared with existing results [38,40,41], this article eliminates the potential assumption restrictions of ideal mode information and mode synchronization in the derivation of the procedure, thus providing broader applicability and more design freedom.

摘要

•A resilient-DETS is designed to more efficiently reduce the releasing frequency of sampled signals transmitted through the communication network, saving limited communication resources while resisting the impact of aperiodic DoS jamming attacks.•Based on HMM with PAHI and switched system analysis techniques, a unified framework is established by fully considering the effects of DETS, aperiodic DoS jamming attacks, partially unknown TRs and CPs, and sufficient conditions are developed to guarantee that the resulting switched residual model is stochastically stable and has a weighted dissipativity performance. In such a way, the resource awareness, resilience performance, and dissipativity performance can be analyzed simultaneously. Besides, a quantitative relation among the system parameters, sampling period, DoS duration, sleeping period, and dissipativity performance index is explicitly described in the stability criterion.•A co-design criterion of asynchronous FDF and resilient-DETS for MJSs with PAHI is deduced, which allows the unknown elements to be existing in the TRM of Markov chain, the CPM of observed signal, or in both of them, and even in the extreme scenario where the TPs and CPs are completely unknown, the proposed method can still provide feasible solutions while existing methods cannot [7–11]. Compared with existing results [38,40,41], this article eliminates the potential assumption restrictions of ideal mode information and mode synchronization in the derivation of the procedure, thus providing broader applicability and more design freedom.