Optimization Problems in Model-Based Diagnosis

The critical step in model-based fault detection and diagnosis is the design of residuals. Residuals have to provide for decoupling from disturbances, enhancement for fault isolation and ideal conditions for statistical testing. When these requirements cannot be satisfied exactly, optimal (approxima...

Full description

Saved in:
Bibliographic Details
Published inIFAC Proceedings Volumes Vol. 36; no. 8; pp. 1 - 8
Main Authors Gertler, Janos, Hu, Yongtong
Format Journal Article
LanguageEnglish
Published 01.06.2003
Subjects
disturbance decoupling
enhanced residuals
Fault diagnosis
residual generation
statistical testing
Fault diagnosis
enhanced residuals
residual generation
disturbance decoupling
statistical testing
Online AccessGet full text

Cover

More Information
Summary:The critical step in model-based fault detection and diagnosis is the design of residuals. Residuals have to provide for decoupling from disturbances, enhancement for fault isolation and ideal conditions for statistical testing. When these requirements cannot be satisfied exactly, optimal (approximate) solutions are sought. Three of these optimal approaches are reviewed in the paper and some of their limitations pointed out.
ISSN:1474-6670
DOI:10.1016/S1474-6670(17)35751-8