Uncertainty-aware fusion of probabilistic classifiers for improved transformer diagnostics

Jose Ignacio Aizpurua, Victoria M. Catterson, Brian G. Stewart, Stephen D. J. McArthur, Brandon Lambert, James G. Cross

Research output: Contribution to journalArticle

Abstract

Transformers are critical assets for the reliable operation of the power grid. Transformers may fail in service if monitoring models do not identify degraded conditions in time. Dissolved gas analysis (DGA) focuses on the examination of dissolved gasses in transformer oil to diagnose the state of a transformer. Fusion of black-box classifiers, also known as an ensemble of diagnostics models, have been used to improve the accuracy of diagnostics models across many fields. When independent classifiers diagnose the same fault, this method can increase the veracity of the diagnostics. However, if these methods give conflicting results, it is not always clear which model is most accurate due to their black-box nature. In this context, the use of white-box models can help resolve conflicted samples effectively by incorporating uncertainty information and improve the classification accuracy. This paper presents an uncertainty aware fusion method to combine black-box and white-box diagnostics methods. The effectiveness of the proposed approach is validated using two publicly available DGA datasets.
LanguageEnglish
Pages1-13
Number of pages13
JournalIEEE Transactions on Systems Man and Cybernetics: Systems
Early online date3 Dec 2018
DOIs
Publication statusE-pub ahead of print - 3 Dec 2018

Fingerprint

Classifiers
Fusion reactions
Gas fuel analysis
Insulating oil
Uncertainty
Monitoring

Keywords

  • condition monitoring
  • transformer diagnostics
  • ensembles
  • classifiers
  • uncertainty

Cite this

@article{09547de61ac847af9f11083a0c47a651,
title = "Uncertainty-aware fusion of probabilistic classifiers for improved transformer diagnostics",
abstract = "Transformers are critical assets for the reliable operation of the power grid. Transformers may fail in service if monitoring models do not identify degraded conditions in time. Dissolved gas analysis (DGA) focuses on the examination of dissolved gasses in transformer oil to diagnose the state of a transformer. Fusion of black-box classifiers, also known as an ensemble of diagnostics models, have been used to improve the accuracy of diagnostics models across many fields. When independent classifiers diagnose the same fault, this method can increase the veracity of the diagnostics. However, if these methods give conflicting results, it is not always clear which model is most accurate due to their black-box nature. In this context, the use of white-box models can help resolve conflicted samples effectively by incorporating uncertainty information and improve the classification accuracy. This paper presents an uncertainty aware fusion method to combine black-box and white-box diagnostics methods. The effectiveness of the proposed approach is validated using two publicly available DGA datasets.",
keywords = "condition monitoring, transformer diagnostics, ensembles, classifiers, uncertainty",
author = "Aizpurua, {Jose Ignacio} and Catterson, {Victoria M.} and Stewart, {Brian G.} and McArthur, {Stephen D. J.} and Brandon Lambert and Cross, {James G.}",
note = "{\circledC} 2018 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.",
year = "2018",
month = "12",
day = "3",
doi = "10.1109/TSMC.2018.2880930",
language = "English",
pages = "1--13",
journal = "IEEE Transactions on Systems Man and Cybernetics: Systems",
issn = "2168-2216",
publisher = "IEEE Advancing Technology for Humanity",

}

Uncertainty-aware fusion of probabilistic classifiers for improved transformer diagnostics. / Aizpurua, Jose Ignacio; Catterson, Victoria M.; Stewart, Brian G.; McArthur, Stephen D. J.; Lambert, Brandon; Cross, James G.

In: IEEE Transactions on Systems Man and Cybernetics: Systems, 03.12.2018, p. 1-13.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Uncertainty-aware fusion of probabilistic classifiers for improved transformer diagnostics

AU - Aizpurua, Jose Ignacio

AU - Catterson, Victoria M.

AU - Stewart, Brian G.

AU - McArthur, Stephen D. J.

AU - Lambert, Brandon

AU - Cross, James G.

N1 - © 2018 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.

PY - 2018/12/3

Y1 - 2018/12/3

N2 - Transformers are critical assets for the reliable operation of the power grid. Transformers may fail in service if monitoring models do not identify degraded conditions in time. Dissolved gas analysis (DGA) focuses on the examination of dissolved gasses in transformer oil to diagnose the state of a transformer. Fusion of black-box classifiers, also known as an ensemble of diagnostics models, have been used to improve the accuracy of diagnostics models across many fields. When independent classifiers diagnose the same fault, this method can increase the veracity of the diagnostics. However, if these methods give conflicting results, it is not always clear which model is most accurate due to their black-box nature. In this context, the use of white-box models can help resolve conflicted samples effectively by incorporating uncertainty information and improve the classification accuracy. This paper presents an uncertainty aware fusion method to combine black-box and white-box diagnostics methods. The effectiveness of the proposed approach is validated using two publicly available DGA datasets.

AB - Transformers are critical assets for the reliable operation of the power grid. Transformers may fail in service if monitoring models do not identify degraded conditions in time. Dissolved gas analysis (DGA) focuses on the examination of dissolved gasses in transformer oil to diagnose the state of a transformer. Fusion of black-box classifiers, also known as an ensemble of diagnostics models, have been used to improve the accuracy of diagnostics models across many fields. When independent classifiers diagnose the same fault, this method can increase the veracity of the diagnostics. However, if these methods give conflicting results, it is not always clear which model is most accurate due to their black-box nature. In this context, the use of white-box models can help resolve conflicted samples effectively by incorporating uncertainty information and improve the classification accuracy. This paper presents an uncertainty aware fusion method to combine black-box and white-box diagnostics methods. The effectiveness of the proposed approach is validated using two publicly available DGA datasets.

KW - condition monitoring

KW - transformer diagnostics

KW - ensembles

KW - classifiers

KW - uncertainty

UR - https://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=6221021

U2 - 10.1109/TSMC.2018.2880930

DO - 10.1109/TSMC.2018.2880930

M3 - Article

SP - 1

EP - 13

JO - IEEE Transactions on Systems Man and Cybernetics: Systems

T2 - IEEE Transactions on Systems Man and Cybernetics: Systems

JF - IEEE Transactions on Systems Man and Cybernetics: Systems

SN - 2168-2216

ER -