Transformer fault diagnosis method based on multiple-level denoising and recursive graph

doi:10.3969/j.issn.2097-0706.2026.02.006

Abstract

Abstract:

To address the challenges of extracting features from transformer vibration signals under noise interference conditions and the low utilization of temporal information，a fault diagnosis method combining multi-level denoising and recurrence plots is proposed. Multi-level denoising of the vibration signals was achieved through complete ensemble empirical mode decomposition with adaptive noise（CEEMDAN）combined with adaptive wavelet thresholding. Recurrence analysis was then used to transform the time-series signals into recurrence plots，preserving temporal dependencies and nonlinear dynamic features. A convolutional neural network（CNN）was employed to leverage its strengths in image processing，using the recurrence plots as inputs to extract spatial features，thereby enabling transformer fault diagnosis. The results show that the proposed method achieves a low false alarm rate and a diagnostic accuracy of 98.125%，demonstrating its effectiveness for diagnosing transformer faults based on vibration signals.

Key words: transformer, fault diagnosis, deep learning, multi-level denoising, complete ensemble empirical mode decomposition with adaptive noise, recurrence analysis, convolutional neural network

CLC Number:

TM411：TP18

ZHANG Ping, WU Zhang, LIU Feng, LIU Feng, LI Jian. Transformer fault diagnosis method based on multiple-level denoising and recursive graph[J]. Integrated Intelligent Energy, 2026, 48(2): 59-67.

Figures/Tables 12

Fig.1

Fig.2

Fig.3

Table 1

Table 2

Fig.4

Fig.5

Fig.6

Table 3

Fig.7

Fig.8

Fig.9

References 25

[1]	汲胜昌, 张凡, 师愉航, 等. 基于振动信号的电力变压器机械状态诊断方法研究综述[J]. 高电压技术, 2020, 46(1): 257-272.
	JI Shengchang, ZHANG Fan, SHI Yuhang, et al. Review on vibration-based mechanical condition monitoring in power transformers[J]. High Voltage Engineering, 2020, 46(1): 257-272.
[2]	张凡, 汲胜昌, 师愉航, 等. 电力变压器绕组振动及传播特性研究[J]. 中国电机工程学报, 2018, 38(9): 2790-2798, 2849.
	ZHANG Fan, JI Shengchang, SHI Yuhang, et al. Research on transformer winding vibration and propagation characteristics[J]. Proceedings of the CSEE, 2018, 38(9): 2790-2798, 2849.
[3]	刘云鹏, 董王英, 许自强, 等. 基于卷积神经网络的变压器套管故障红外图像识别方法[J]. 高压电器, 2021, 57(10): 134-140.
	LIU Yunpeng, DONG Wangying, XU Ziqiang, et al. Infrared image recognition method on fault of transformer bushing based on convolutional neutral networks[J]. High Voltage Apparatus, 2021, 57(10): 134-140.
[4]	GUAN S, YANG H Q, WU T Y. Transformer fault diagnosis method based on TLR-ADASYN balanced dataset[J]. Scientific Reports, 2023, 13: 23010. doi: 10.1038/s41598-023-49901-9
[5]	IEC. IEC60599:Mineraloil-filledelectricalequipmentinservice—Guidanceontheinterpretationofdissolvedandfreegasesanalysis[S]. Geneva,Switzerland:IEC, 2007.
[6]	OUYANG X, ZHOU Q, LUO Z H, et al. Analysis on the magnetic flux leakage distribution in the transformer under different winding deformation and typical working condition[C]// Proceedings of 2020 IEEE International Conference on High Voltage Engineering and Application (ICHVE). IEEE, 2020: 1-4.
[7]	BEREGI S, TAKACS D, STEPAN G. Bifurcation analysis of wheel shimmy with non-smooth effects and time delay in the tyre-ground contact[J]. Nonlinear Dynamics, 2019, 98(1): 841-858. doi: 10.1007/s11071-019-05123-1
[8]	曾浩政, 殷林飞. 基于NSMFO-BERT算法的电力系统多目标优化经济调度研究[J]. 综合智慧能源, 2025, 47(4): 98-106. doi: 10.3969/j.issn.2097-0706.2025.04.008
	ZENG Haozheng, YIN Linfei. Research on multi-objective optimal economic dispatch of power systems based on NSMFO-BERT algorithm[J]. Integrated Intelligent Energy, 2025, 47(4): 98-106. doi: 10.3969/j.issn.2097-0706.2025.04.008
[9]	孙玉凤, 介石磊, 王明. 基于自适应阈值小波的发动机摩擦声发射信号降噪[J]. 机械设计与研究, 2024, 40(3): 181-184, 188.
	SUN Yufeng, JIE Shilei, WANG Ming, Noise reduction of engine friction acoustic emission signal based on adaptive threshold wavelet[J]. Machine Design & Research, 2024, 40(3): 181-184, 188.
[10]	朱仁杰, 宋恩哲, 姚崇, 等. 强噪声背景下基于CEEMDAN与BRECAN的船舶电机故障诊断[J]. 中国舰船研究, 2025, 20(2): 20-29.
	ZHU Renjie, SONG Enzhe, YAO Chong, et al. Marine motor fault diagnosis based on CEEMDAN and BRECAN under strong noise conditions[J]. Chinese Journal of Ship Research, 2025, 20(2): 20-29.
[11]	刘建锋, 范一凡, 宋伊宁, 等. 基于改进连续变分模态分解和深度残差网络及漏磁信号的变压器绕组故障诊断[J/OL]. 电网技术,1-12(2025-04-14)[2025-04-28].https://doi.org/10.13335/j.1000-3673.pst.2024.1262.
	LIU Jianfeng, FAN Yifan, SONG Yining, et al. Fault diagnosis of transformer windings based on improved sequential variational mode decomposition, deep residual network and magnetic leakage signals[J/OL]. Power System Technology, 1-12(2025-04-14)[2025-04-28].https://doi.org/10.13335/j.1000-3673.pst.2024.1262.
[12]	杨义, 李晓华, 李俊聪, 等. 基于SSWT-GLCM与改进WOA-SVM的变压器机械故障时频诊断[J]. 振动、测试与诊断, 2024, 44(6): 1135-1143, 1247.
	YANG Yi, LI Xiaohua, LI Juncong, et al. Time frequency diagnosis of transformer mechanical fault based on SSWT-GLCM and improved WOA-SVM[J]. Journal of Vibration, Measurement & Diagnosis, 2024, 44(6): 1135-1143, 1247.
[13]	郝旺身, 李继康, 董辛旻, 等. 基于声信号递归Hilbert变换的轴承故障诊断研究[J]. 机床与液压, 2024, 52(4): 195-199.
	HAO Wangshen, LI Jikang, DONG Xinmin, et al. Research on bearing fault diagnosis based on acoustic signal recursive Hilbert transform[J]. Machine Tool & Hydraulics, 2024, 52(4): 195-199.
[14]	施保华, 吴婷, 赵子睿. 基于递归图和增强残差网络的轴承故障诊断[J]. 轴承, 2024(12): 87-94.
	SHI Baohua, WU Ting, ZHAO Zirui. Fault diagnosis for bearings based on recurrence plot and enhance ResNet[J]. Bearing, 2024(12): 87-94.
[15]	赵晓震, 吴歌, 王建, 等. 基于频响重构和数字图像处理的变压器绕组变形故障诊断方法[J/OL]. 高电压技术,1-12(2024-09-12)[2025-02-20].https://doi.org/10.13336/j.1003-6520.hve.20240694.
	ZHAO Xiaozhen, WU Ge, WANG Jian, et al. Fault diagnosis method of transformer winding deformation based on frequency response reconstruction and digital image processing[J/OL]. High Voltage Engineering, 1-12(2024-09-12)[2025-02-20].https://doi.org/10.13336/j.1003-6520.hve.20240694.
[16]	马宏忠, 周宇, 李凯, 等. 基于振动的变压器绕组压紧状态评估方法[J]. 电力系统自动化, 2015, 39(18): 127-132.
	MA Hongzhong, ZHOU Yu, LI Kai, et al. Evaluation method for transformer windings pressed state based on vibration[J]. Automation of Electric Power Systems, 2015, 39(18): 127-132.
[17]	石鑫, 朱永利, 宁晓光, 等. 基于深度自编码网络的电力变压器故障诊断[J]. 电力自动化设备, 2016, 36(5): 122-126.
	SHI Xin, ZHU Yongli, NING Xiaoguang, et al. Transformer fault diagnosis based on deep auto-encoder network[J]. Electric Power Automation Equipment, 2016, 36(5): 122-126.
[18]	张凯, 王金袖, 杨雪峰, 等. 基于二次EMD分解的风电场混合储能系统配置优化[J/OL]. 综合智慧能源,1-7(2025-01-06)[2025-02-20]. http://kns.cnki.net/kcms/detail/41.1461.TK.20250103.1424.008.html.
	ZHANG Kai, WANG Jinxiu, YANG Xiufeng, et al. Optimization of hybrid energy storage system configuration for a wind power plant basedon dual EMD decomposition[J/OL]. Integrated Intelligent Energy,1-7(2025-01-06)[2025-02-20]. http://kns.cnki.net/kcms/detail/41.1461.TK.20250103.1424.008.html.
[19]	沈景涛, 武哲, 张强, 等. 基于WATD-MTF与改进的残差网络齿轮箱故障诊断研究[J]. 振动与冲击, 2025, 44(7): 247-257.
	SHEN Jingtao, WU Zhe, ZHANG Qiang, et al. Gearbox fault diagnosis based on WATD-MTF and improved residual network[J]. Journal of Vibration and Shock, 2025, 44(7): 247-257.
[20]	王海军, 王涛, 俞慈君. 基于递归量化分析的CFRP超声检测缺陷识别方法[J]. 浙江大学学报(工学版), 2024, 58(8): 1604-1617.
	WANG Haijun, WANG Tao, YU Cijun. CFRP ultrasonic detection defect identification method based on recursive quantitative analysis[J]. Journal of Zhejiang University (Engineering Science), 2024, 58(8): 1604-1617.
[21]	CHENG X Y, MA G S, WU Z Y, et al. Automatic defect depth estimation for ultrasonic testing in carbon fiber reinforced composites using deep learning[J]. NDT & E International, 2023, 135: 102804. doi: 10.1016/j.ndteint.2023.102804
[22]	谢丽蓉, 严侣, 吐松江·卡日, 等. 基于特征判定系数的电力变压器振动信号故障诊断[J]. 电力工程技术, 2024, 43(3): 217-225.
	XIE Lirong, YAN Lyu, TUSONGJIANG Kari, et al. Power transformer vibration signal fault diagnosis based on feature determination coefficient[J]. Electric Power Engineering Technology, 2024, 43(3): 217-225
[23]	LEE J C, PARK D H, JUNG H S, et al. Design for carbon fiber lamination of PMI foam cored CFRP sandwich composite applied to automotive rear spoiler[J]. Fibers and Polymers, 2020, 21(1): 156-161. doi: 10.1007/s12221-020-9489-4
[24]	赵莉华, 徐立, 刘艳, 等. 基于点对称变换与图像匹配的变压器机械故障诊断方法[J]. 电工技术学报, 2021, 36(17): 3614-3626.
	ZHAO Lihua, XU Li, LIU Yan, et al. Transformer mechanical fault diagnosis method based on symmetrized dot patter and image matching[J]. Transactions of China Electrotechnical Society, 2021, 36(17): 3614-3626.
[25]	荆新岚, 黄民, 马超. 基于可视图谱特征融合的行星齿轮箱故障诊断[J]. 电子测量技术, 2024, 47(24): 171-178.
	JING Xinlan, HUANG Min, MA Chao. Fault diagnosis of planetary gearbox based on visual spectral feature fusion[J]. Electronic Measurement Technology, 2024, 47(24): 171-178.

状态	SNR/dB	RMSE	CE
无故障	23.7~26.2	0.008 6~0.023 1	0.88~0.94
绕组变形	19.8~23.1	0.013 4~0.041 5	0.87~0.96
绕组松动	20.8~24.2	0.009 9~0.034 7	0.85~0.93
铁芯松动	21.5~24.6	0.007 2~0.030 8	0.88~0.93

信号	SNR/dB	RMSE	CE	准确率/%
多重降噪	20.8~26.2	0.007 2~0.041 5	0.85~0.96	98.125
CEEMDAN	17.8~22.5	0.015 7~0.062 0	0.82~0.96	93.438
自适应小波阈值	16.5~20.7	0.019 9~0.066 9	0.83~0.91	91.250

标签	精确率/%	召回率/%	F₁分数/%	样本数
1	100.00	100.00	100.00	80
2	96.34	98.75	97.53	80
3	96.30	97.50	96.89	80
4	100.00	96.25	98.09	80