A multi strategy improved sparrow search algorithm (ISSA) based on sample equalization and fused feature optimization was proposed to optimize the eXtreme gradient boosting (XGBoost) transformer fault diagnosis model for the problem of low diagnostic accuracy caused by imbalanced fault samples and difficult effective feature extraction in oil immersed transformers. Initially, the conditional tabular GAN (CTGAN) was employed to balance the dissolved gas analysis (DGA) dataset. Subsequently, a 23-dimensional candidate set of transformer fault features was established using the correlation ratio method. The optimal feature set for input was then determined by combining the Spearman correlation coefficient with the feature importance derived from the XGBoost model. Furthermore, the SSA was enhanced by incorporating improved Henon_Circle chaotic mapping, the fish eagle algorithm, the differential evolution algorithm, the Cauchy distribution, and an elite opposition-based learning strategy. The superiority of the ISSA was validated through performance tests using four typical test functions. Ultimately, the ISSA-XGBoost diagnostic model was constructed, and its diagnostic performance was verified through multiple comparative experiments. The results show that the proposed method achieves a diagnostic accuracy of 97.8%, which can effectively improve the sample classification capability and is more suitable for transformer fault diagnosis with imbalanced samples.
WANG Chen, ZHAO Jia-qi, XU Xiao-nan, WANG Shu-xia, WU Rong-rong.
ISSA-XGBoost Transformer Fault Diagnosis Based on Sample Equalization and Feature Optimization[J].
Science Technology and Engineering, 2026, 26(27): 11763-11775 DOI:10.12404/j.issn.1671-1815.2504631
ZhuLi, WangXiaohao, LiHao, et al. Transformer fault diagnosis method based on variation sparrow search algorithm and improved smote under unbalanced samples[J]. High Voltage Engineering, 2023, 49(12): 4993-5001.
LiuYunpeng, HeJiahui, XuZiqiang, et al. Equalization method of power transformer fault sample based on SVM SMOTE[J]. High Voltage Engineering, 2020, 46(7): 2522-2529.
[10]
DouzasG, BacaoF, LastF. Improving imbalanced learning through a heuristic oversampling method based on K-means and SMOTE[J]. Information Sciences, 2018, 465: 1-26.
[11]
MajzoubA H, ElgedawyI, AkaydınÖ, et al. HCAB-SMOTE: a hybrid clustered affinitive borderline SMOTE approach for imbalanced data binary classification[J]. Arabian Journal for Science and Engineering, 2020, 45(4): 3205-3222.
[12]
HeH, BaiY, GarciaE A, et al. ADASYN: adaptive synthetic sampling approach for imbalanced learning[C]// 2008 IEEE International Joint Conference on Neural Networks. Hong Kong: IEEE, 2008: 1322-1328.
[13]
XuL, SkoularidouM, Cuesta-infanteA, et al. Modeling tabular data using conditional GAN[J]. Advances in Neural Information Processing Systems, 2019, 32: 7335-7345.
ZhouXiaohua, FengYuchen, ChenLei, et al. Transformer fault diagnosis based on SVM optimized by the improved bald eagle search algorithm[J]. Power System Protection and Control, 2023, 51(8): 118-126.
BaoJinshan, YangDingkun, ZhangJing, et al. Transformer fault diagnosis method based on feature extraction and INGO-SVM[J]. Power System Protection and Control, 2024, 52(7): 24-32.
ZhangKao, HeKailin, YangPeihao. Research on power transformer fault diagnosis algorithm based on fuzzy reinforcement learning[J]. Integrated Intelligent Energy, 2024, 46(10): 48-55.
KangJiayu, ZhangShenxi, ZhangQingping, et al. Fault diagnosis method of transformer based on ANOVA and BO-SVM[J]. High Voltage Engineering, 2023, 49(5): 1882-1891.
LiangFazheng, ZhangLian, YangJiahao, et al. CatBoost-SHAP transformer fault diagnosis study for unbalanced data[J]. Hunan Electric Power, 2024, 44(5): 131-138.
WangYongli, ZhaoZhonghua, ZhangYinuo, et al. Material demand forecasting for distribution networks based on improved particle swarm algorithm and extreme learning machine modeling[J]. Science Technology and Engineering, 2025, 25(15): 6410-6418.
LiangXiangyu, HuYelin, MaXiangyang, et al. Fault diagnosis of rolling bearings based on feature mode decomposition and multiscale fuzzy dispersion entropy[J]. Science Technology and Engineering, 2025, 25(1): 176-185.
LiaoCaibo, YangJinxin, HuXiong, et al. Fault diagnosis method for transformers based on feature selection of dissolved gas in oil[J]. Electric Power Engineering Technology, 2024, 43(1): 192-200.
[32]
GoodfellowI, Pouget-abadieJ, MirzaM, et al. Generative adversarial nets[C]// Advances in Neural Information Processing Systems. Montreal: Curran Associates, 2014: 2672-2680.
[33]
XuL. Synthesizing tabular data using generative adversarial networks[D]. Cambridge: Massachusetts Institute of Technology, 2018.
ZhangGuozhi, ChenKang, FangRongxing, et al. Transformer fault diagnosis method based on DGA and whale algorithm optimizing Logit Boost-decision tree[J]. Power System Protection and Control, 2023, 51(7): 63-72.
MinYongzhi, LingShiwen, WangGuo. Transformer fault diagnosis based on hybrid feature selection and IOMA-CNN[J]. Power System Protection and Control, 2024, 52(23): 1-9.
[40]
ChenT, GuestrinC. XGBoost: a scalable tree boosting system[C]// Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York: ACM, 2016: 1-10.
[41]
XueJ, ShenB. A novel swarm intelligence optimization approach: sparrow search algorithm[J]. Systems Science & Control Engineering, 2020, 8(1): 22-43.
[42]
DehghaniM, TrojovskP. Osprey optimization algorithm: a new bio-inspired metaheuristic algorithm for solving engineering optimization problems[J]. Frontiers in Mechanical Engineering, 2023, 8: 1126450.
[43]
DraaA, BouzoubiaS, BoukhalfaI. A sinusoidal differential evolution algorithm for numerical optimization[J]. Applied Soft Computing, 2015, 27: 99-126.
[44]
WangM, WangJ S, LiX D, et al. Harris hawk optimization algorithm based on Cauchy distribution inverse cumulative function and tangent flight operator[J]. Applied Intelligence, 2022 (10): 10999-11026.
RaoU M, FofanaI, RajeshK N, et al. Identification and application of machine learning algorithms for transformer dissolved gas analysis[J]. IEEE Transactions on Dielectrics and Electrical Insulation, 2021, 28(5): 1828-1835.
LiLei, WangTingtao, HeJianan, et al. Transformer fault diagnosis strategy considering parameter optimization of oversampler and classifier[J]. Electric Power Automation Equipment, 2023, 43(1): 209-217.
YinJinliang. Study on oil immersed power transformer fault diagnosis based on relevance vector machine[D]. Baoding: North China Electric Power University, 2013.
[53]
DaiJ, SongH, ShengG, et al. Dissolved gas analysis of insulating oil for power transformer fault diagnosis with deep belief network[J]. IEEE Transactions on Dielectrics and Electrical Insulation, 2017, 24(5): 2828-2835.