Designing an event-triggered [formula omitted] filter with possibly inconsistent modes for Markov jump systems

Abstract

The design of event-triggered H∞ filters for Markov jump systems is studied. The transition rates of the Markov chain allow for being uncertain and even totally unknown. Meanwhile, the plant considered and its filter are allowed to have possibly inconsistent modes, resulting in a hidden Markov model (HMM). A dual-threshold periodic event-triggered scheme is introduced to save network channel resources. Under the event-triggered scheme, a sufficient condition of mean-square exponential stability and H∞ disturbance-attenuation performance is proposed for the HMM-based filtering error system. Then, a design approach for the desired event-triggered H∞ filter is developed by eliminating potential nonlinearities, without limiting the row sum of uncertain items in the transition rate matrix to zero. Finally, two examples, including a tunnel diode circuit, are provided to verify the effectiveness of the design approach.