Structural Analysis for Diagnosis - the Matching Problem Revisited

Roozbeh Izadi-Zamanabadi, Mogens Blanke

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

Abstract

Aiming at design of algorithms for fault diagnosis, structural analysis of systems offers concise yet easy overall analysis. Graph-based matching, which is the essential technique to obtain redundant information for diagnosis, is re-considered in this paper. Matching is re-formulated as a problem of relating faults to known parameters and measurements of a system. Using explicit fault modelling, minimal over-determined subsystems are shown to provide necessary redundancy relations from the matching. Details of the method are presented and a realistic example used to clearly describe individual steps.
Original languageEnglish
Title of host publicationProceedings of the 15. IFAC World Congress
PublisherElsevier
Publication date2002
DOIs
Publication statusPublished - 2002
EventProceedings of the 15. IFAC World Congress - Barcelona, Spain
Duration: 1 Jan 2002 → …
Conference number: 15

Conference

ConferenceProceedings of the 15. IFAC World Congress
Number15
CityBarcelona, Spain
Period01/01/2002 → …

Keywords

  • FDI
  • Fault diagnosis
  • Structural analysis
  • Autonomous systems

Cite this

Izadi-Zamanabadi, R., & Blanke, M. (2002). Structural Analysis for Diagnosis - the Matching Problem Revisited. In Proceedings of the 15. IFAC World Congress Elsevier. https://doi.org/10.3182/20020721-6-ES-1901.00792