|
ROBUST FAULT DETECTION BY LFR MODELS
Francesco A. Cuzzola* Sergio Bittanti**
* Danieli Automation S.p.A. - Viale Fulvio Testi 124 -
I-20092 Cinisello Balsamo, Italy. e-mail:
cuzzola@aut.ee.ethz.ch
** Politecnico di Milano - Dipartimento di Elettronica e
Informazione, Via G. Ponzio 34/5 - I-20133 Milano, Italy.
In this work we propose a technique for Fault Detection based on Linear Fractional Representations (LFR) and on Linear Matrix Inequality theory (LMI). The technique is more general with respect to some classical techniques based on the assumption that the fault is acting on the considered process as an input. More precisely, the fault is represented as a time varying parameter influencing the dynamics of the considered process.
Keywords: Fault Detection, Linear Matrix Inequalities, Linear Fractional Representations
Session slot T-Tu-M12: Uncertainty & Robustness in FDI/Area code 7e : Fault Detection, Supervision and Safety of Technical Processes
 | 