Transformer fault diagnosis is crucial for ensuring the safe and stable operation of power systems. However, traditional diagnostic methods suffer from issues such as insufficient feature utilization and poor model interpretability. To address these challenges, this paper proposes a hybrid fault diagnosis model. First, the model utilizes SHapley Additive exPlanations values for feature selection. Then, the improved Harris Hawks optimization is proposed to optimize the hyperparameters of eXtreme gradient boosting model (Xgboost), further improving the model’s classification performance. Experimental results demonstrate that the proposed model outperforms existing methods across multiple performance indicators, achieving an accuracy of 0.9509, which represents a 2.68% improvement compared to the Xgboost model. This model fully leverages the advantages of each method, effectively solving problems like insufficient feature utilization, poor model interpretability, and limited performance of optimization algorithms in traditional fault diagnosis methods. It provides a more reliable and efficient solution for transformer fault diagnosis.
YaoHXuYGuoQ, et al.Study on transformer fault diagnosisbased on improved deep residual shrinkage network and optimized residual variational autoencoder. Energy Rep2025; 13: 1608–1619.
2.
NanfakAEkeSKomCH, et al.A dissolved gases analysis method for power transformer faults diagnosis based on the observation of subsets of labelled fault data. J Electr Eng Technol2025; 20(4): 2019–2028.
3.
BitneyVDUlyanovVYAnMM. The need for using various interpretation methods for power transformer DGA results. In: 2023 5th international youth conference on Radio Electronics, Electrical and Power Engineering (REEPE), Moscow, Russian Federation, 16–18 March 2023, pp. 1–6.
4.
MuhamadNAPhungBTBlackburnTR, et al.Comparative study and analysis of DGA methods for transformer mineral oil. In: 2007 IEEE Lausanne power tech, Lausanne, Switzerland, 01–05 July 2007, pp. 45–50.
5.
ManojTRangaCGhoneimSSM, et al.Alternate and effective dissolved gas interpretation to understand the transformer incipient faults. IEEE Trans Dielectr Electr Insul2023; 30: 1231–1239.
6.
WajidARehmanAUIqbalS, et al.Comparative performance study of Dissolved Gas Analysis (DGA) methods for identification of faults in power transformer. Int J Energy Res2023; 2023: 1–14.
7.
TanXQiJGanJQ, et al.Multi-filter semi-supervised transformer model for fault diagnosis. Eng Appl Artif Intell2023; 124: 106498.
8.
TusongjiangKDuLAisikaerR, et al.Fault diagnosis of power transformer based on improved ACGAN under imbalanced data. Comput Mater Contin2023; 75: 4573–4592.
9.
ZhangXNingH. Transformer fault diagnosis method based on relieff and HPO-SVM. University Politehnica of Bucharest Scientific Bulletin Series C - Electri Eng Comput Sci2023; 85: 137–152.
10.
RaoSZouGYangS, et al.A feature selection and ensemble learning based methodology for transformer fault diagnosis. Appl Soft Comput2024; 150: 111072.
11.
LiuJCaiBYanS, et al.Transformer fault diagnosis based on the improved QPSO and random forest. Meas Sci Technol2024; 35: 096206.
12.
ShangHLiuZWeiY, et al.A novel Fault Diagnosis method for a power transformer based on multi-scale approximate entropy and optimized convolutional networks. Entropy2024; 26: 186.
13.
LuWShiCFuH, et al.A power Transformer Fault Diagnosis method based on improved sand cat swarm optimization algorithm and bidirectional gated recurrent unit. Electronics2023; 12: 672.
14.
MalarvizhiDPrakashA. An extensive analysis of feature selection in machine learning. Int J Res Appl Sci Eng Technol2023; 11(12): 1197–1202.
15.
KhanSNoorSJavedT, et al.XGBoost-enhanced ensemble model using discriminative hybrid features for the prediction of sumoylation sites. BioData Min2025; 18(1): 12.
16.
KhanSAlQahtaniSANoorS, et al.Deep learning based intelligent model for prediction of sumoylation sites using discriminative features. BMC Bioinf2024; 25(1): 284.
17.
NoorSNaseemAAwanHH, et al.Deep-m5U: a deep learning-based approach for RNA 5-methyluridine modification prediction using optimized feature integration. BMC Bioinf2024; 25(1): 360.
18.
ShahidMRazaAAkbarS, et al.pACPs-DNN: predicting anticancer peptides using novel peptide transformation into evolutionary and structure matrix-based images with self-attention deep learning model. Comput Biol Chem2025; 117: 108441.
19.
AkbarSRazaAAwanHH, et al.pNPs-CapsNet: predicting neuropeptides using protein language models and FastText encoding-based weighted multi-view feature integration with deep capsule neural network. ACS Omega2025; 10(12): 12403–12416.
20.
DingCYuDLiuX, et al.Research on Transformer Fault Diagnosis by WOA-SVM based on feature selection and data balancing. IEEJ Transactions Elec Engng2025; 20(1): 41–49.
21.
BhitkarABoharaSKulkarniT, et al.A comprehensive analysis of optimization algorithms for large-scale machine learning. Int J Sci Res Eng Manag2024; 8(11): 1–5.
22.
AbualigahLElazizMASumariP, et al.Reptile Search Algorithm (RSA): a nature-inspired meta-heuristic optimizer. Expert Syst Appl2022; 191: 116158.
23.
PołapDWoźniakM. Red fox optimization algorithm. Expert Syst Appl2021; 166: 114107.
24.
TasicAJovanovicLBacaninN, et al.Towards sustainable societies: convolutional neural networks optimized by modified crayfish optimization algorithm aided by AdaBoost and XGBoost for waste classification tasks. Appl Soft Comput2025; 175: 113086.
25.
ShafiullahM,AbidoMAand Al-Mohammed AH. Metaheuristic optimization techniques. In: AndraeA (ed) Power system fault diagnosis. Amsterdam: Elsevier, 2022, pp. 27–68.
26.
GharehchopoghFSGhafouriSNamaziM, et al.Advances in manta ray foraging optimization: a comprehensive survey. J Bionic Eng2024; 21(2): 953–990.
27.
BashaJBacaninNVukobratN, et al.Chaotic harris hawks optimization with quasi-reflection-based learning: an application to enhance CNN design. Sensors (Peterb, NH)2021; 21(19): 6654.
28.
ZivkovicMBacaninNArandjelovicJ, et al.Novel harris hawks optimization and deep neural network approach for intrusion detection. In: UddinMSJamwalPKBansalJC (eds). Proceedings of international joint conference on advances in computational intelligence. Singapore: Springer Nature Singapore, 2022, pp. 239–250.
29.
BacaninNVukobratNZivkovicM, et al.Improved harris hawks optimization adapted for artificial neural network training. In: KahramanCCebiSCevik OnarS, et al. (eds). Intelligent and fuzzy techniques for emerging conditions and digital transformation. Berlin: Springer International Publishing, 2022, pp. 281–289.
30.
HernesMAdaszyńskiJTutakP. Credit risk modeling using interpreted XGBoost. European Management Studies2023; 2023(3): 46–70.
31.
MaulanaAFaisalFRNoviandyTR, et al.Machine learning approach for diabetes detection using fine-tuned XGBoost algorithm. Infolitika J Data Sci2023; 1(1): 1–7.
32.
DanLShiCZhangJ, et al.Application of Xgboost algorithm based on machine learning in reservoir prediction of offshore oilfield. In: LinJ (ed). Proceedings of the international field exploration and development conference 2020. Singapore: Springer, 2021, pp. 3157–3167.
33.
YuanRDingSWangC. Comparative analysis of machine-learning models for recognizing lane-change intention using vehicle trajectory data. Infrastructure2023; 8(11): 156.
34.
DobrojevicMJovanovicLBabicL, et al.Cyberbullying sexism harassment identification by metaheurustics-tuned eXtreme gradient boosting. Comput Mater Contin2024; 80(3): 4997–5027.
35.
HeidariAAMirjaliliSFarisH, et al.Harris hawks optimization: algorithm and applications. Future Gener Comput Syst2019; 97: 849–872.
36.
LawnikM. Generalized logistic map and its application in chaos based cryptography. J Phys: Conf Ser2017; 936: 012017.
37.
LiYZhaoYShangY, et al.An improved firefly algorithm with dynamic self-adaptive adjustment. PLoS One2021; 16(7): e0255951.
38.
YueXJiangLZengY, et al.Whale optimization algorithm using pinhole imaging reverse learning and golden sine strategy. Int J Cognit Inf Nat Intell2024; 18(1): 1–16.
39.
MirjaliliSMirjaliliSMLewisA. Grey wolf optimizer. Adv Eng Software2014; 69: 46–61.
40.
LuoSWangBGaoQ, et al.Stacking integration algorithm based on CNN-BiLSTM-Attention with XGBoost for short-term electricity load forecasting. Energy Rep2024; 12: 2676–2689.
41.
SunXXuZFengY, et al.RBC inventory-management system based on XGBoost model. Indian J Hematol Blood Transfus2021; 37(1): 126–133.
42.
BabaeiGGiudiciP. InstanceSHAP: an instance-based estimation approach for Shapley values. SSRN Journal2022, [Epub ahead of print].
43.
SerenoGSRauberTWMunaroCJ. Specialized classifiers with SHAP feature selection: application to the Tennessee Eastman process. In: 2023 15th IEEE International Conference on Industry Applications (INDUSCON). São Bernardo do Campo, Brazil, 22–24 November 2023, pp. 1338–1345.
44.
McHughML. Interrater reliability: the kappa statistic. Biochem Med2012; 22(3): 276–282.
45.
ZhouXTianTLiuN, et al.Research on gas production law of free gas in oil-immersed power transformer under discharge fault of different severity. Front Energy Res2023; 10: 1056604.
46.
NugrahaWSasongkoA. Hyperparameter tuning on classification algorithm with grid search. SISTEMASI2022; 11(1): 391.
47.
KaleAW. Evolutionary optimization of deep learning models: enhance performance through hyperparameter and structural tuning with evolutionary algorithms. Adv Nonlinear Var Inequalities2024; 27(3): 632–649.
48.
ZangXCaoZXiaoM, et al.Research on online monitoring of grid current transformer based on transformers and BiGRU. J Comput Methods Sci Eng2023; 23(3): 1137–1149.
49.
NoorSAwanHHHashmiAS, et al.Optimizing performance of parallel computing platforms for large-scale genome data analysis. Computing (U K)2025; 107(86): 86.
