Finite‐horizon fault estimation for time‐varying systems with multiple fading measurements under torus‐event–based protocols

• Published in: International journal of robust and nonlinear control Vol. 29; no. 13; pp. 4594 - 4608
• Main Authors: Ju, Yamei, Wei, Guoliang, Ding, Derui, Liu, Shuai
• Format: Journal Article
• Language: English
• Published: Bognor Regis Wiley Subscription Services, Inc 10.09.2019
• Subjects: Computer simulation; Difference equations; Discrete time systems; Fading; H-infinity control; Horizon; H∞ fault estimation; randomly occurring faults; recursive Riccati difference equations (RDEs); Schedules; Toruses; torus‐event–based protocols
• ISSN: 1049-8923; 1099-1239
• DOI: 10.1002/rnc.4640

Summary In this paper, the issue of the finite‐horizon H∞ fault estimation is dealt with for a class of discrete time‐varying systems subject to randomly occurring faults and multiple fading measurements. The missing phenomena may occur in a random way from different sensors, which is represented by an individual stochastic variable meeting a certain probability distribution. Furthermore, in order to alleviate the communication burden, the torus‐event–based protocols are adopted to schedule the data transmissions only when some significant events occur. Our aim of the presented issue is to estimate the fault such that, with multiple fading measurements via the received information governed by torus‐event–based protocols, the H∞ index is satisfied over a given finite horizon. Sufficient conditions are obtained for the desired time‐varying estimator in terms of the technique of stochastic analysis and the methods of completing squares. The desired estimator gains are calculated by working out two backward recursive Riccati difference equations. Finally, a numerical simulation is given to verify the usefulness of our designed fault estimation approach.