This paper addresses the low diagnostic accuracy of dissolved gas analysis (DGA) in transformer faults by proposing an improved honey pot algorithm (IHBA) to optimize the deep mixed kernel extreme learning machine (DHKELM). First, the dissolved gas data in transformer oil is preprocessed to reduce the discrepancies in the magnitude of fault data. Kernel Principal Component Analysis (KPCA) is then applied to reduce dimensionality and extract effective features, thus decreasing the correlation among the data.Next, traditional honey pot algorithm (HBA) is improved by introducing Cubic chaotic mapping, random value perturbation strategies, elite tangent search, and differential mutation strategies. The performance of IHBA is tested using three typical benchmark functions, demonstrating that IHBA possesses stronger stability and optimization capabilities. IHBA is then utilized to optimize the parameters of DHKELM, establishing the IHBA-DHKELM model for transformer fault diagnosis. Finally, the features extracted via KPCA are used as the input set for the model and compared with various transformer fault diagnosis models. Simulation results indicate that IHBA-DHKELM achieves higher diagnostic accuracy for transformer faults.

Changshun Lv, Ziwei Zhu, Yinuo Deng. Transformer Fault Diagnosis Based on KPCA-IHBA-DHKELM. International Journal of Big Data Intelligent Technology (2024), Vol. 5, Issue 1: 137-152. https://doi.org/10.38007/IJBDIT.2024.050115.

[1] Cao, Y.. Research on Transformer Fault Diagnosis Method Based on Improved Grey Wolf Algorithm Optimized BP Neural Network. Electrical Switch2024;62(02), 82-85..

[2] Y. Zhang, L. Zhao, J, "Forecasting of Dissolved Gases in Power Transformer Oil Based on DOG -LSSVM Regression and Artificial Bee Colony," 2018 International Conference on Power System Technology (POWERCON), Guangzhou, China, 2018, pp. 3620-3625.

[3] Yuwei Zhang. Transformer fault diagnosis based on Fuzzy rogers four ratio method. Electrotech Technol 2021;(12):89–92.

[4] Yi Liu, Yuanping Ni. Transformer fault diagnosis method based on three-ratio gray correlation analysis. High Volt Technol2002;28(10):16–7.

[5] Chengming Zhang, Jufang Xie, Song Yu, Chao Tang, Dong Hu. An improved fault diagnosis method for transformer Duval pentagon1 based on spatial analysis theory. High Volt Technol 2022;1–11

[6] Xiao, Y.; Pan, W.; Guo, X.; Bi, S.; Feng, D.; Lin, S. Fault Diagnosis of Traction Transformer Based on Bayesian Network. Energies 2020, 13, 4966.

[7] Z. Zhan, J. Chen, W. Chen, "Transformer Fault Diagnosis Method Based on Fuzzy Logic and D-S Evidence Theory," 2022 5th International Conference on Energy, Electrical and Power Engineering (CEEPE), Chongqing, China, 2022, pp. 470-475.

[8] Jinglong Jia, Tao Yu, Zijie Wu et al. 2017 Fault diagnosis method of transformer based on convolutional neural network[J] Electrical Measurement & Instrumentation 54 62-67

[9] Yang Fang Ming, Liu Chuan, Sun Yong et al. 2014 Fault Prediction Based on Dissolved Gas Concentration from Insulating Oil in Power Transformer Using Neural Network[J] 2789 312-317

[10] Meng W.D. 2020 Transformer Fault Diagnosis Based on BP Neural Network J. Communication power technology 37 84-86

[11] Yang X, Pang S, Shen W, et al. Aero engine fault diagnosis using an optimized extreme learning machine[J]. International Journal of Aerospace Engineering, 2016, 2016(1): 7892875.

[12] Hu, Beibei, and Yunhe Cheng. "Predicting regional carbon price in China based on multi-factor HKELM by combining secondary decomposition and ensemble learning." Plos one 18.12 (2023): e0285311.

[13] Wang W, Cui X, Qi Y, et al. Prediction model of coal seam gas content based on kernel principal component analysis and IDBO-DHKELM[J]. Measurement Science and Technology, 2024, 35(11): 115113.

[14] Guerbas, F., Benmahamed, Y., Teguar, Y. et al. Neural networks and particle swarm for transformer oil diagnosis by dissolved gas analysis. Sci Rep 14, 9271 (2024).

[15] H. Peng and C. Zhao, "Research on fault diagnosis of KPCA-WOA-BP transformer," 2024 5th International Conference on Computer Engineering and Application (ICCEA), Hangzhou, China, 2024, pp. 1487-1493.

[16] M. Zhang and W. Chen, "Fault Diagnosis of Power Transformer Based on SSA—MDS Pretreatment," in IEEE Access, vol. 10, pp. 92505-92515, 2022.

[17] Leifeng He and Ying Huang 2022 A transformer fault diagnosis method based on grey wolf optimization algorithm optimized support vector machine[J] Hongshuihe 41 84-88

[18] Guomin Xie and Jialiang Wang . "Transformer Fault Identification Method Based on Hybrid Sampling and IHBA-SVM." Journal of Electronic Measurement and Instrumentation, vol. 36, no. 12, 2022, pp. 77-85.

[19] Xuan Chen. "Research on Transformer Feature Selection Method Based on Grey Wolf Algorithm." Electrical Materials, vol. 2024, no. 04, 2024, pp. 90-92, 96. DOI: 10.16786/j.cnki.1671-8887.eem.2024.04.024.

[20] Xin Zheng and Chun Shi et al. "Fault Diagnosis Method for Coal Mine Transformers Based on ISSA-SVM." Electromechanical Engineering Technology, vol. 51, no. 07, 2022, pp. 31-34, 49

[21] Xiaoqin Zhang and Chunqiang Hu. "Data Acquisition and Monitoring System Attack Detection Model Based on Improved Extreme Learning Machine." Journal of Nanjing University of Aeronautics and Astronautics, 2021, pp. 708-717

[22] Liqun Shang,Yadong Hou, et al. "Transformer Fault Diagnosis Based on IDOA-DHKELM." High Voltage Engineering, vol. 49, no. 11, 2023, pp. 4726-4735. DOI: 10.13336/j.1003-6520.hve.20221483.

[23] Jing Yan, Xueying Zhang, et al. "Regression Prediction Model Combining Stack-based Supervised AE and Variable Weight ELM." Computer Engineering, vol. 48, no. 08, 2022, pp. 62-69, 76.

[24] Fatma A. Hashim, Essam H. Houssein,et al. Honey Badger Algorithm: New metaheuristic algorithm for solving optimization problems,Mathematics and Computers in Simulation,Volume 192,2022,Pages 84-110,ISSN 0378-4754

[25] Lu W, Shi C, Fu H, et al. Fault diagnosis method for power transformers based on improved golden jackal optimization algorithm and random configuration network[J]. IEEE Access, 2023, 11: 35336-35351.

[26] Y. Wu, Y. Zhang, X. Zhong and L. Cheng, "A Power Transformer Fault Diagnosis Method-Based Hybrid Improved Seagull Optimization Algorithm and Support Vector Machine," in IEEE Access, vol. 10, pp. 17268-17286, 2022.

[27] Zhang, X.; Sun, Z. Application of Improved PNN in Transformer Fault Diagnosis. Processes 2023, 11, 474. 

[28] L. Kou, C. Liu, G. -w. Cai, Z. Zhang, X. -j. Li and Q. -d. Yuan, "Fault Diagnosis for Power Converters based on Random Forests and Feature Transformation," 2020 IEEE 9th International Power Electronics and Motion Control Conference (IPEMC2020-ECCE Asia), Nanjing, China, 2020, pp. 1821-1826, doi: 10.1109/IPEMC-ECCEAsia48364.2020.9367970. 

[29] Pengfei Cao and Yongping Gan,. "Transformer Fault Diagnosis Based on Vector Weighting Algorithm Optimized ELM." Environmental Technology, vol. 41, no. 10, 2023, pp. 136-142.

[30] Jiang J, Liu Z, Wang P, et al. Improved Crow Search Algorithm and XGBoost for Transformer Fault Diagnosis[C]//Journal of Physics: Conference Series. IOP Publishing, 2023, 2666(1): 012040. [4] BRAND K P,KOPAINSKY J. Particle densities in a decaying SF6 plasm[J]. Applied Physics,1978,16(4):425-432.