Dynamic Event‐Based Resilient Filtering for Discrete‐Time Singular Markov Jump Cpss Against Hybrid Cyber‐Attacks

ABSTRACT This paper investigates the mixed and passive filtering problem for discrete‐time singular Markov jump cyber‐physical systems (CPSs) with a dynamic event‐triggered policy (ETP) under hybrid cyber‐attacks. To conserve network bandwidth, a dynamic ETP is proposed and adopted. Owing to asynchronous phenomena between the modes of singular Markov jump CPSs and the designed filter, a hidden Markov model is utilized for the filter design. The filtering error systems are converted into time‐delayed singular Markov jump CPSs with partially known conditional probabilities. By leveraging a series of linear matrix inequalities (LMIs), the criteria of stochastic admissibility for the filtering error system are derived. Moreover, the co‐design of the resilient filter and dynamic ETP is proposed. Finally, two case studies are conducted to verify the correctness and validity of the proposed scheme.