Fault Diagnosis of Dissolved Gas Analysis for Transformer Based on XGBoost

Chuqi Yang

doi:10.54254/2753-8818/2025.AD26526

Theoretical and Natural ScienceOpen access

Fault Diagnosis of Dissolved Gas Analysis for Transformer Based on XGBoost

Research Article

Open Access

Fault Diagnosis of Dissolved Gas Analysis for Transformer Based on XGBoost

Chuqi Yang ^1*

¹ Nanjing University

^*Corresponding author: 2242306675@qq.com

Published on 3 September 2025

TNS Vol.134

ISSN (Print): 2753-8826

ISSN (Online): 2753-8818

ISBN (Print): 978-1-80590-343-7

ISBN (Online): 978-1-80590-344-4

Download Cover

Abstract

Aiming at the fault diagnosis problem of oil-immersed transformers, this paper proposes a transformer oil dissolved gas analysis (DGA) fault diagnosis method based on the XGBoost algorithm. Traditional diagnostic methods have defects such as vague boundary values and poor interpretability. However, the XGBoost algorithm can effectively capture the nonlinear relationship between DGA data and fault types by iteratively optimizing the additive model. In the study, 1260 sets of DGA data were preprocessed, including mean normalization to eliminate the influence of dimensions, construction of gas ratio features to enhance sensitivity, and conversion of fault types into numerical labels. The model was optimized by setting hyperparameters such as learning_rate and max_ depth, and using 5-fold cross-validation and early stopping mechanism. Experimental results show that the accuracy of the XGBoost model on the test set reaches 93.6%, which is significantly higher than that of LSTM (82.3%) and PSO-LSTM (85.7%), and the RMSE and MAE indicators are better. The research shows that this method can accurately diagnose transformer faults and provide effective technical support for the safe operation of power systems.

Keywords:

Transformer fault diagnosis, DGA, XGBoost algorithm, Gradient boosting decision tree, Hyperparameter setting

View PDF

References

[1]. Wang Hao. Common Faults and Handling Methods of Oil-immersed Transformers [J]. Shandong Industrial Technology, 2018, (22): 184.

[2]. Lin Beimin. Case Analysis of Typical Transformer Faults Based on the Three-ratio Method [J]. Electrical Technology, 2023, 24(10): 63-67.

[3]. Hu Daofu, Wen Shanshan, He Yiming. Transformer Fault Diagnosis Based on BP Neural Network and Its Application [J]. Journal of Electric Power Science and Technology, 2008, (02): 72-75.

[4]. Hong Yitian, Chen Weihua, Chen Tongzhu, et al. Operation Fault Diagnosis Method for Large Power Transformers Based on XGBOOST Algorithm [J]. Electrical Engineering, 2024, (11): 56-59.

[5]. Ruan Yi, Zhang Haotian, Sun Jian, et al. A NRBO-XGBoost Transformer Fault Diagnosis Method Based on DGA [J]. Journal of Chaohu University, 2024, 26(06): 87-93+128.

[6]. Jia Haoyang, Qian Yu. Transformer Fault Diagnosis Based on Bayesian Optimized XGBoost Algorithm [J]. Journal of Yellow River Conservancy Technical Institute, 2023, 35(02): 37-43.

[7]. Li Yiming, Liang Zhiqing, Xu Mian. Malicious DGA Domain Name Detection Based on LSTM and Attention Mechanism [J]. Network Security Technology and Application, 2024, (12): 32-34.

[8]. Chen Zesheng, Zhou Min, Feng Lichun, et al. Malicious Domain Name Recognition of DGA Based on XGBoost and Particle Swarm Optimization Algorithm [J]. Journal on Communications, 2024, 45(S2): 27-32

[9]. Wang Mingyang, Ma Xuejun, Ge Lijuan, et al. Research on Transformer Fault Diagnosis Based on KOA-CNN-LSTM [J]. Journal of Inner Mongolia Agricultural University (Natural Science Edition), 2025, 46(04): 65-73

[10]. Jia Rubin, Zhang Yajun, Tian Feng, et al. Application of Convolutional Neural Network in Fault Diagnosis of Oil-immersed Transformers [J]. Electrical Engineering, 2024, (10): 89-93.