kth.sePublications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Fault Diagnostics of Power Transformers Using Autoencoders and Gated Recurrent Units with Applications for Sensor Failures
KTH, School of Electrical Engineering and Computer Science (EECS), Electrical Engineering, Electric Power and Energy Systems.ORCID iD: 0000-0002-6428-2241
KTH, School of Electrical Engineering and Computer Science (EECS), Electrical Engineering, Electric Power and Energy Systems.ORCID iD: 0000-0003-4763-9429
2022 (English)In: 2022 17th International Conference on Probabilistic Methods Applied to Power Systems (PMAPS)., Institute of Electrical and Electronics Engineers (IEEE), 2022Conference paper, Published paper (Refereed)
Abstract [en]

With the industrial internet of things and SG technology developing, more and more operation data could be accessible, and condition-based maintenance shows promise for electrical equipment. This paper aims to develop a data-driven fault diagnosis utilizing operation data for high voltage equipment condition monitoring. To understand the asset management of power transformers, an interview is conducted as the expertise input for the study. The paper uses deep learning in an unsupervised way to model normal behaviors and identify underlying operational risks. The autoencoders are used to compress the raw data and extract the key features and the gated recurrent unit to model the dependencies between normal behaviors of power transformers. Finally, the method employs control charts to generate the alarm to indicate the underlying anomalies. The paper uses an online dataset to test the applications for sensor failures. The results show that the method can identify the operational risks before sensor failures.

Place, publisher, year, edition, pages
Institute of Electrical and Electronics Engineers (IEEE), 2022.
Series
International Conference on Probabilistic Methods Applied to Power Systems, ISSN 2642-6730
Keywords [en]
asset management, condition monitoring, fault diagnostics, machine learning, power transformers
National Category
Other Electrical Engineering, Electronic Engineering, Information Engineering
Identifiers
URN: urn:nbn:se:kth:diva-320411DOI: 10.1109/PMAPS53380.2022.9810620ISI: 000853744900063Scopus ID: 2-s2.0-85135003993OAI: oai:DiVA.org:kth-320411DiVA, id: diva2:1708981
Conference
17th International Conference on Probabilistic Methods Applied to Power Systems (PMAPS), JUN 12-15, 2022, Manchester, England
Note

QC 20221107

Part of proceedings: ISBN 978-1-6654-1211-7

Available from: 2022-11-07 Created: 2022-11-07 Last updated: 2024-03-18Bibliographically approved

Open Access in DiVA

No full text in DiVA

Other links

Publisher's full textScopus

Authority records

Cui, YueBertling Tjernberg, Lina

Search in DiVA

By author/editor
Cui, YueBertling Tjernberg, Lina
By organisation
Electric Power and Energy Systems
Other Electrical Engineering, Electronic Engineering, Information Engineering

Search outside of DiVA

GoogleGoogle Scholar

doi
urn-nbn

Altmetric score

doi
urn-nbn
Total: 39 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf