Open AccessResearch on Transformer State Evaluation Method Based on Fault Feature Extraction
Author(s) -
Shuang Liu,
Rui Han,
Yongtao Jin,
Wenhao Wang,
Haofan Lin,
Zhi Yang
Publication year - 2020
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1639/1/012028
Subject(s) - transformer , reliability engineering , logistic regression , computer science , engineering , machine learning , electrical engineering , voltage
In order to realize the dynamic evaluation of transformer status and gradually improve the level of equipment fault diagnosis, a transformer condition assessment method based on fault feature extraction was proposed in this paper. Based on the data of the key state quantity of dissolved gas in transformer oil, this method extracted the transformer fault characteristics through various mathematical methods such as hypothesis testing and logistic language regression, and selected new indexes that could better reflect the operating state of transformer, so as to judge the operating state of transformer. In this paper, a large number of indicators were screened, and the regression method was used to find the indicators to judge the fault, without other external conditions. The analysis results showed that the average prediction accuracy of the proposed method was more than 95% under multiple cross validation, which had high engineering application value.