Study of application of composite sampling and improved LightGBM algorithm to the diagnosis of unbalanced transformer fault samples

Shanshan Liao; Dongsheng He; Yangyang Xie; Kaining Qin; Jingmin Fan; Shungui Liu; Zhiping Ou

doi:10.1117/12.2611471

14 October 2021 Study of application of composite sampling and improved LightGBM algorithm to the diagnosis of unbalanced transformer fault samples

Shanshan Liao, Dongsheng He, Yangyang Xie, Kaining Qin, Jingmin Fan, Shungui Liu, Zhiping Ou

Author Affiliations +

Proceedings Volume 11930, International Conference on Mechanical Engineering, Measurement Control, and Instrumentation; 119302N (2021) https://doi.org/10.1117/12.2611471
Event: International Conference on Mechanical Engineering, Measurement Control, and Instrumentation (MEMCI 2021), 2021, Guangzhou, China

Abstract

A new diagnosis method for unbalanced transformer fault samples was proposed in this study to solve the problems of low accuracy and long training time existing in the traditional transformer state evaluation and diagnosis algorithms under un-balanced sample distribution. First, the improved SMOTE alg-orithm was used to collect the minority among the unbalanced transformer fault samples, which were then normalized. Next, the ENN algorithm was adopted to solve the data noise problem generated after sampling, followed by the under-sampling pro-cessing, and the samples of different fault types were made to reach the balanced state through the composite sampling. In the end, the overfitting problem was solved through the improved LightGBM algorithm, and the balanced transformer fault sam-ples were diagnosed. The experimental results showed that under unbalanced samples, the diagnostic accuracy rate of the proposed method reached 95.1%, which was 24.7% higher than that reached by other traditional fault diagnosis methods. Moreover, the proposed method improved the indexes like recall rate, F1-score value and precision ratio by 20%, thus solving the low fault diagnosis accuracy and long training time of traditional artificial intelligence (AI) algorithms under unbalanced data distribution of transformer fault samples.

Citation Download Citation

Shanshan Liao, Dongsheng He, Yangyang Xie, Kaining Qin, Jingmin Fan, Shungui Liu, and Zhiping Ou "Study of application of composite sampling and improved LightGBM algorithm to the diagnosis of unbalanced transformer fault samples", Proc. SPIE 11930, International Conference on Mechanical Engineering, Measurement Control, and Instrumentation, 119302N (14 October 2021); https://doi.org/10.1117/12.2611471

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available

Members: $17.00

Non-members: $21.00 ADD TO CART

PROCEEDINGS
12 PAGES

DOWNLOAD PAPER SAVE TO MY LIBRARY

GET CITATION

RIGHTS & PERMISSIONS

Get copyright permission Get copyright permission on Copyright Marketplace

KEYWORDS

Transformers

Composites

Statistical modeling

Data modeling

Gases

Detection and tracking algorithms

Diagnostics

Show All Keywords

Keywords/Phrases

Search In:

Publication Years