Event-based passive filtering for Markov jump singularly perturbed complex networks

This paper studies the event-based passive filtering issue for a series of time-delayed complex networks in the discrete-time domain. These networks feature singularly perturbed and Markov jump parameters, where the Markov chain models abrupt changes in node couplings and structural parameters, while the singularly perturbed parameter addresses discrepancies in time scales. Considering the limited communication bandwidth resource, a dynamic event-triggered mechanism is applied. This paper aims to design a reliable filter to estimate the states of complex networks, ensuring the filtering error system’s stochastic stability and achieving an expected passive performance. Using convex optimization techniques and Lyapunov methodology, we derive sufficient criteria to ensure the stability of the filtering error system and the existence of such a filter. To validate the feasibility of the proposed method, both numerical and a practical examples are presented in the simulation part.