Robust Design of Fault Isolation Observers

• Published in: Automatica (Oxford) Vol. 34; no. 11; pp. 1421 - 1429
• Main Authors: SHEN, LI-CHENG, HSU, PAU-LO
• Format: Journal Article
• Language: English
• Published: Oxford Elsevier Ltd 01.11.1998; Elsevier
• Subjects: Applied sciences; Computer science; control theory; systems; Control theory. Systems; eigenstructure assignment; engine diagnosis; Exact sciences and technology; fault isolation; Miscellaneous; Unknown inputs; unstructured uncertainties; eigenstructure assignment; Unknown inputs; unstructured uncertainties; fault isolation; engine diagnosis; Insulation; Uncertain system; Linear system; Observer; Failure detection; Robustness; Discrete system
• ISSN: 0005-1098; 1873-2836
• DOI: 10.1016/S0005-1098(98)00087-9

The robust diagnostics design has gained more attentions recently in real applications. It is known that the freedom in eigenstructure assignment leads to the well-conditioned design which is robust to unstructured uncertainties (Shen et al., 1995, 1998). Moreover, the unknown input fault detection observer (UIFDO) decouples the known inputs, the structured uncertainties, from the diagnostic residuals. In this paper, we present an unknown input fault isolation observer (UIFIO) based on the UIFDO by using the eigenstructure assignment. It achieves robust fault isolation subject to both structured and unstructured uncertainties. Moreover, the optimal gain matrix of the present diagnostic observer is obtained to improve its robustness without degrading diagnostic performance. Experimental results of an automotive engine show the validity and applicability of the proposed robust UIFIO design.