Premium
Robust Unknown Input Observer Design for Linear Uncertain Time Delay Systems with Application to Fault Detection
Author(s) -
Ahmadizadeh Saeed,
Zarei Jafar,
Karimi Hamid Reza
Publication year - 2014
Publication title -
asian journal of control
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.769
H-Index - 53
eISSN - 1934-6093
pISSN - 1561-8625
DOI - 10.1002/asjc.765
Subject(s) - control theory (sociology) , fault detection and isolation , residual , observer (physics) , sensitivity (control systems) , linear system , norm (philosophy) , computer science , fault (geology) , signal (programming language) , filter (signal processing) , noise (video) , algorithm , mathematics , engineering , electronic engineering , artificial intelligence , control (management) , actuator , mathematical analysis , physics , quantum mechanics , seismology , law , political science , image (mathematics) , computer vision , programming language , geology
In this paper, a novel approach is proposed to design a robust fault detection observer for uncertain linear time delay systems. The system is composed of both norm‐bounded uncertainties and exogenous signals (noise, disturbance, and fault) which are considered to be unknown. The main contribution of this paper is to present unknown input observer ( UIO )‐based fault detection system which shows the maximum sensitivity to fault signals and the minimum sensitivity to other signals. Since the system contains uncertainty terms, an H ∞ model‐matching approach is used in design procedure. The reference residual signal generator system is designed so that the fault signal has maximum sensitivity while the exogenous signals have minimum sensitivity on the residual signal. Then, the fault detection system is designed by minimizing the estimation error between the reference residual signal and the UIO residual signal in the sense of H ∞ norm. A sufficient condition for the existence of such a filter is exploited in terms of certain linear matrix inequalities ( LMIs ). Application of the proposed method in a numerical example and an engineering process are simulated to demonstrate the effectiveness of the proposed algorithm. Simulation results show the validity of the proposed approach to detect the occurrence of faults in the presence of modeling errors, disturbances, and noise.