Condition monitoring and failure prognosis of IGBT inverters based on on-line characterization

Andrew Babel, Annette Muetze, Roland Seebacher, Klaus Krischan, Elias G. Strangas

Publikation: Beitrag in Buch/Bericht/KonferenzbandBeitrag in einem Konferenzband

Abstract

An inverter characterization technique is used to improve inverter fault diagnosis and prognosis techniques with the demonstrative bond wire lift-off fault. Knowledge of the nonlinear voltage-current characteristic of each inverter device is used to achieve these improvements. Three contributions are made in this paper to the field of inverter condition monitoring by using the characterization technique: use of the inverter device characteristics to improve a classifier-based diagnostic technique, direct use of the device characteristics for diagnosis, and improving the calculation of remaining useful life with the device characteristics. The diagnostic method improvements are verified experimentally by adding resistances to create an artificial fault. Thermal and electrical simulations are used to demonstrate the improvements made to the prognostic method.

Originalspracheenglisch
Titel2014 IEEE Energy Conversion Congress and Exposition, ECCE 2014
Herausgeber (Verlag)Institute of Electrical and Electronics Engineers
Seiten3059-3066
Seitenumfang8
ISBN (elektronisch)9781479956982
DOIs
PublikationsstatusVeröffentlicht - 11 Nov 2014
VeranstaltungIEEE Energy Conversion Congress and Exposition - , USA / Vereinigte Staaten
Dauer: 14 Sep 201418 Sep 2014

Konferenz

KonferenzIEEE Energy Conversion Congress and Exposition
LandUSA / Vereinigte Staaten
Zeitraum14/09/1418/09/14

ASJC Scopus subject areas

  • !!Fuel Technology
  • !!Energy Engineering and Power Technology

Fields of Expertise

  • Sustainable Systems

Treatment code (Nähere Zuordnung)

  • Theoretical
  • Experimental

Fingerprint

Untersuchen Sie die Forschungsthemen von „Condition monitoring and failure prognosis of IGBT inverters based on on-line characterization“. Zusammen bilden sie einen einzigartigen Fingerprint.

Dieses zitieren