Robust fault hiding approach for T–S fuzzy systems with unmeasured premise variables

• Published in: Information sciences Vol. 589; pp. 690 - 715
• Main Authors: Quadros, Mariella Maia, Leite, Valter J.S., Palhares, Reinaldo Martínez
• Format: Journal Article
• Language: English
• Published: Elsevier Inc 01-04-2022
• Subjects: Fault tolerant control; Reconfiguration block; T–S fuzzy systems; Virtual actuator; Virtual sensor; Virtual actuator; T–S fuzzy systems; Reconfiguration block; Virtual sensor; Fault tolerant control
• ISSN: 0020-0255; 1872-6291
• DOI: 10.1016/j.ins.2021.12.129

This paper proposes the design of a robust Takagi–Sugeno (T–S) fuzzy reconfiguration block for systems with unmeasured premise variables. The reconfiguration block is composed of a virtual sensor and a virtual actuator whose design is obtained by means of new conditions based on Linear Matrix Inequalities (LMIs) with guaranteed H∞ performance. The main contribution of the new proposed fault hiding approach concerning the current methodologies in the literature is to deal with systems subject to sensor and actuator faults with unmeasured premise variables, including the cases in which the measurements of the premise variables are affected by sensor faults. Besides, using the polytopic representation of the faults, the reconfiguration block design provides to the system robustness to a set of different sensor and actuator faults that may occur (including total and partial faults of different magnitudes). A real-time experiment consisting of two coupled tanks described as a T–S fuzzy model subject to sensor and actuator faults, with unmeasured premise variable, illustrates the performance of the proposed methodology. The experiment shows that the proposed approach can compensate the different types of faults, even when the measurement of one of the premise variables is affected by the fault.