Finite-horizon fault estimation under imperfect measurements and stochastic communication protocol: Dealing with finite-time boundedness

Abstract

In this paper, we analyze the finite-horizon fault estimation issue for a kind of time-varying nonlinear systems with imperfect measurement signals under the stochastic communication protocol (SCP). The imperfect measurements result from randomly occurring sensor nonlinearities obeying sensor-wise Bernoulli distributions. The Markov-chain-driven SCP is introduced to regulate the signal transmission to alleviate the communication congestion. The aim of the considered issue is to propose the design algorithm of a group of time-varying fault estimators such that the estimation error dynamics satisfies both the H∞ and the finite-time boundedness (FTB) performance requirements. First, sufficient conditions are set up to guarantee the existence of the satisfactory H∞ FTB fault estimators through intensive stochastic analyses and matrix operations. Then, the gains of such fault estimators are explicitly parameterized by resorting to the solution to recursive linear matrix inequalities. Finally, the correctness of the devised fault estimation approach is demonstrated by a numerical example.