Power Transformer Failure Prediction: Classification in Imbalanced Time Series
This paper describes a study on applying data mining techniques to power transformer failure prediction. The data set used consisted not only on DGA tests, but also in other tests done to the transformer’s insulating oil. This dataset presented several challenges, such as highly imbalanced classes (...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Universidade do Porto
2017-09-01
|
| Series: | U.Porto Journal of Engineering |
| Subjects: | |
| Online Access: | https://journalengineering.fe.up.pt/article/view/87 |
| _version_ | 1818543866833272832 |
|---|---|
| author | Eduardo e Oliveira Vera L. Miguéis Luís Guimarães José Borges |
| author_facet | Eduardo e Oliveira Vera L. Miguéis Luís Guimarães José Borges |
| author_sort | Eduardo e Oliveira |
| collection | DOAJ |
| description | This paper describes a study on applying data mining techniques to power transformer failure prediction. The data set used consisted not only on DGA tests, but also in other tests done to the transformer’s insulating oil. This dataset presented several challenges, such as highly imbalanced classes (common in failure prediction problems), and the temporal nature of the observations.
To overcome these challenges, several techniques were applied for prediction and better understand the dataset. Pre-processing and temporality incorporation in the dataset is discussed. For prediction, a 1-class and 2-class SVM, decision trees and random forests, as well as a LSTM neural network were applied to the dataset.
As the prediction performance was low (high false-positive rate), we conducted a test to ascertain if the amount of data collected was sufficient. Results indicate that the frequency of data collection was not adequate, hinting that the degradation period was shorter than the periodicity of data collection. |
| first_indexed | 2024-12-11T22:40:59Z |
| format | Article |
| id | doaj.art-13ef24aa622343e5a244a72a8dc106e4 |
| institution | Directory Open Access Journal |
| issn | 2183-6493 |
| language | English |
| last_indexed | 2024-12-11T22:40:59Z |
| publishDate | 2017-09-01 |
| publisher | Universidade do Porto |
| record_format | Article |
| series | U.Porto Journal of Engineering |
| spelling | doaj.art-13ef24aa622343e5a244a72a8dc106e42022-12-22T00:47:47ZengUniversidade do PortoU.Porto Journal of Engineering2183-64932017-09-0132344810.24840/2183-6493_003.002_000487Power Transformer Failure Prediction: Classification in Imbalanced Time SeriesEduardo e Oliveira0Vera L. Miguéis1Luís Guimarães2José Borges3INESC TEC; Universidade do PortoINESC TEC; Universidade do PortoINESC TEC; Universidade do PortoINESC TEC; Universidade do PortoThis paper describes a study on applying data mining techniques to power transformer failure prediction. The data set used consisted not only on DGA tests, but also in other tests done to the transformer’s insulating oil. This dataset presented several challenges, such as highly imbalanced classes (common in failure prediction problems), and the temporal nature of the observations. To overcome these challenges, several techniques were applied for prediction and better understand the dataset. Pre-processing and temporality incorporation in the dataset is discussed. For prediction, a 1-class and 2-class SVM, decision trees and random forests, as well as a LSTM neural network were applied to the dataset. As the prediction performance was low (high false-positive rate), we conducted a test to ascertain if the amount of data collected was sufficient. Results indicate that the frequency of data collection was not adequate, hinting that the degradation period was shorter than the periodicity of data collection.https://journalengineering.fe.up.pt/article/view/87Data MiningFailure PredictionTime SeriesPower Transformer |
| spellingShingle | Eduardo e Oliveira Vera L. Miguéis Luís Guimarães José Borges Power Transformer Failure Prediction: Classification in Imbalanced Time Series U.Porto Journal of Engineering Data Mining Failure Prediction Time Series Power Transformer |
| title | Power Transformer Failure Prediction: Classification in Imbalanced Time Series |
| title_full | Power Transformer Failure Prediction: Classification in Imbalanced Time Series |
| title_fullStr | Power Transformer Failure Prediction: Classification in Imbalanced Time Series |
| title_full_unstemmed | Power Transformer Failure Prediction: Classification in Imbalanced Time Series |
| title_short | Power Transformer Failure Prediction: Classification in Imbalanced Time Series |
| title_sort | power transformer failure prediction classification in imbalanced time series |
| topic | Data Mining Failure Prediction Time Series Power Transformer |
| url | https://journalengineering.fe.up.pt/article/view/87 |
| work_keys_str_mv | AT eduardoeoliveira powertransformerfailurepredictionclassificationinimbalancedtimeseries AT veralmigueis powertransformerfailurepredictionclassificationinimbalancedtimeseries AT luisguimaraes powertransformerfailurepredictionclassificationinimbalancedtimeseries AT joseborges powertransformerfailurepredictionclassificationinimbalancedtimeseries |