Fault diagnosis method integrating dual-threshold preliminary screening and edge-level dynamic graph convolution

doi:10.11772/j.issn.1001-9081.2025091114

Journal of Computer Applications ›› 2026, Vol. 46 ›› Issue (10): 3400-3406.DOI: 10.11772/j.issn.1001-9081.2025091114

• Frontier and comprehensive applications • Previous Articles

Fault diagnosis method integrating dual-threshold preliminary screening and edge-level dynamic graph convolution

Wanxiong MIN¹, Yicheng YAN¹, Jiuyin WANG¹, Yukuan LIANG¹, Wenyang LIANG¹, Zuozhou PAN², Yong LIAO³()

^1.Goupitan Power Plant，Guizhou Wujiang Hydropower Development Company Limited，Zunyi Guizhou 564408，China
^2.College of Metrology Measurement and Instrument，China Jiliang University，Hangzhou Zhejiang 310018，China
^3.School of Microelectronics and Communication Engineering，Chongqing University，Chongqing 401331，China

Received:2025-09-25 Revised:2025-10-17 Accepted:2025-10-20 Online:2025-11-12 Published:2026-10-10
Contact: Yong LIAO
About author:MIN Wanxiong， born in 1988， senior engineer. His research interests include primary electrical equipment， safety production management.
YAN Yicheng， born in 1994， engineer. His research interests include primary electrical equipment， production management.
WANG Jiuyin， born in 1986， engineer. His research interests include production management and digitalization.
LIANG Yukuan， born in 1998， assistant engineer. His research interests include primary electrical equipment.
LIANG Wenyang， born in 2001， assistant engineer. His research interests include primary electrical equipment.
PAN Zuozhou， born in 1994， Ph. D.， associate professor. His research interests include intelligent diagnosis.
Supported by:
Youth Program of Zhejiang Provincial Natural Science Foundation(LQN25E050004);China Huadian Group Technology Project(GPTDC/2024-1209)

融合双阈值初筛与边级动态图卷积的故障诊断方法

闵万雄¹, 严意成¹, 王久银¹, 梁禹宽¹, 梁文杨¹, 潘作舟², 廖勇³()

^1.贵州乌江水电开发有限责任公司构皮滩发电厂，贵州遵义 564408
^2.中国计量大学计量测试与仪器学院，杭州 310018
^3.重庆大学微电子与通信工程学院，重庆 401331

通讯作者: 廖勇
作者简介:闵万雄（1988—），男，贵州遵义人，高级工程师，主要研究方向：电气一次、安全生产管理
严意成（1994—），男，贵州印江人，工程师，主要研究方向：电气一次、生产管理
王久银（1986—），男，贵州遵义人，工程师，主要研究方向：生产管理及其数字化
梁禹宽（1998—），男，贵州贵阳人，助理工程师，主要研究方向：电气一次
梁文杨（2001—），男，贵州遵义人，助理工程师，主要研究方向：电气一次
潘作舟（1994—），男，安徽巢湖人，副教授，博士，主要研究方向：智能诊断
基金资助:
浙江省自然科学基金青年科学基金资助项目(LQN25E050004);中国华电集团科技项目(GPTDC/2024-1209)

Abstract

Abstract:

During operation， power transformers emit vibration signals continuously， which contain many pulses and fluctuations caused by mechanical faults. These signals serve as the primary data source for assessing the operating condition of power transformers. To address the issues of low diagnostic accuracy， poor real-time performance， and inability to handle multi-fault scenarios effectively in the traditional power transformer fault diagnosis techniques， the voiceprint characteristics of internal vibration acoustic signals in power transformers were analyzed， and a fault diagnosis method integrating dual-threshold preliminary screening and edge-level dynamic graph convolution was constructed on the basis of acoustic features. First， a normal/fault signal classification module was proposed to assess the real-time operating condition of power transformers using acoustic short-time energy and zero-crossing rate. Second， an Edge-Level Dynamic Graph Convolutional Network （ELDGCN） was designed to learn spatial correlations among voiceprint signals adaptively by optimizing edge weights. Finally， experiments were conducted under real-world power transformer scenarios. The results show that the proposed method improves the fault identification efficiency significantly， with the identification rate for datasets containing mixed faults reaching 98.17%.

Key words: power transformer, acoustic short-time energy, zero-crossing rate, fault detection, voiceprint signal, Graph Convolutional Network (GCN)

摘要：

电力变压器在运行过程中会持续发出振动信号，其中包含大量由机械故障引起的脉冲和波动，这些信号是评估电力变压器运行状况的主要数据来源。为解决传统电力变压器故障诊断技术诊断精度低、实时性差且无法有效应对多故障场景的问题，通过分析电力变压器内部振动声信号的声纹特征，并基于声学特征构建一个融合双阈值初筛与边级动态图卷积的故障诊断方法。首先，提出一种利用声学短时能量和过零率双阈值评估电力变压器实时运行状况的正常/故障信号分类模块；然后，设计一种边级动态图卷积网络（ELDGCN），通过优化边的权重自适应地学习声纹信号之间的空间相关性。在真实的电力变压器场景下的实验结果表明，采用本文方法显著提升了故障的识别效率，数据集含混合故障的识别率达98.17%。

关键词: 电力变压器, 声学短时能量, 过零率, 故障检测, 声纹信号, 图卷积网络

CLC Number:

TP391.42

Wanxiong MIN, Yicheng YAN, Jiuyin WANG, Yukuan LIANG, Wenyang LIANG, Zuozhou PAN, Yong LIAO. Fault diagnosis method integrating dual-threshold preliminary screening and edge-level dynamic graph convolution[J]. Journal of Computer Applications, 2026, 46(10): 3400-3406.

闵万雄, 严意成, 王久银, 梁禹宽, 梁文杨, 潘作舟, 廖勇. 融合双阈值初筛与边级动态图卷积的故障诊断方法[J]. 《计算机应用》唯一官方网站, 2026, 46(10): 3400-3406.

Figures/Tables 14

References 21

[1]	张宇，时珊珊，张开宇，等. 面向城市轨道交通综合能源站的储能系统鲁棒优化运行策略［J］. 重庆大学学报， 2024， 47（7）： 86-97.
	Zhang Yu， Shi Shanshan， Zhang Kaiyu， et al. Robust optimization operation strategy of energy storage system for integrated energy stations in urban rail transit［J］. Journal of Chongqing University， 2024， 47（7）： 86-97.
[2]	Wu Y， Sun X， Zhang Y， et al. A power transformer fault diagnosis method-based hybrid improved seagull optimization algorithm and support vector machine［J］. IEEE Access， 2022， 10： 17268-17286.
[3]	赵承利，张璐，钟麦英. 基于循环卷积生成对抗网络的风机齿轮箱故障诊断［J］. 山东科技大学学报（自然科学版）， 2024， 43（1）： 109-118.
	Zhao Chengli， Zhang Lu， Zhong Maiying. Fault diagnosis for wind turbine gearbox based on recurrent convolutional generative adversarial network［J］. Journal of Shandong University of Science and Technology （Natural Science）， 2024， 43（1）： 109-118.
[4]	彭宇林，魏静，张爱强，等. 点蚀故障下行星齿轮传动振动特性［J］. 重庆大学学报， 2025， 48（1）： 10-20.
	Peng Yulin， Wei Jing， Zhang Aiqiang， et al. Vibration characteristics of planetary gear transmission under pitting faults［J］. Journal of Chongqing University， 2025， 48（1）： 10-20.
[5]	鲁朋，宋保业，许琳. 基于一维卷积神经网络的多工况轴承故障诊断［J］. 山东科技大学学报（自然科学版）， 2023， 42（5）： 88-96.
	Lu Peng， Song Baoye， Xu Lin. Bearing fault diagnosis in multiple working conditions based on one dimensional convolutional neural network［J］. Journal of Shandong University of Science and Technology （Natural Science）， 2023， 42（5）： 88-96.
[6]	Tenbohlen S， Beura C P， Siegel M. Condition assessment of power transformers by UHF PD measurements［C］// CMD 2022. Piscataway： IEEE， 2022： 5-10.
[7]	Faiz J， Soleimani M. Assessment of computational intelligence and conventional dissolved gas analysis methods for transformer fault diagnosis［J］. IEEE Transactions on Dielectrics and Electrical Insulation， 2018， 25（5）： 1798-1806.
[8]	王丰华，王邵菁，陈颂，等. 基于改进MFCC和VQ的变压器声纹识别模型［J］. 中国电机工程学报， 2017， 37（5）： 1535-1543.
	Wang Fenghua， Wang Shaojing， Chen Song， et al. Voiceprint recognition model of power transformers based on improved MFCC and VQ［J］. Proceedings of the CSEE， 2017， 37（5）： 1535-1543.
[9]	马宏忠，李楠，杨启帆，等. 基于多特征声纹图谱的变压器绕组松动在线故障诊断方法［J］. 电机与控制学报， 2023， 27（5）： 76-87.
	Ma Hongzong， Li Nan， Yang Qifan， et al. On-line fault diagnosis method of transformer winding looseness based on multi-characteristic voiceprint maps［J］. Electric Machines and Control， 2023， 27（5）： 76-87.
[10]	周东旭，王丰华，党晓婧，等. 基于压缩观测与判别字典学习的干式变压器声纹识别［J］. 中国电机工程学报， 2020， 40（19）： 6380-6390.
	Zhou Dongxu， Wang Fenghua， Dang Xiaojing， et al. Dry type transformer voiceprint recognition based on compressed observation and discrimination dictionary learning［J］. Proceedings of the CSEE， 2020， 40（19）： 6380-6390.
[11]	刘云鹏，王博闻，李欢，等. 结合载纤绕组形变测量法的大型变压器绕组多次短路冲击暂态声纹特征［J］. 中国电机工程学报， 2022， 42（1）： 434-447.
	Liu Yunpeng， Wang Bowen， Li Huan， et al. Transient acoustics characteristics of large transformer windings under multiple short-circuit impulse combined with fiber-carrying winding deformation measurement［J］. Proceedings of the CSEE， 2022， 42（1）： 434-447.
[12]	Hassan E， El-Rashidy N， Talaa F M. Review： Mask R-CNN models［J］. Nile Journal of Communication and Computer Science， 2022， 3（1）： 17-27.
[13]	姚万业，冯涛明. 基于改进YOLOv3的变压器定位检测研究［J］. 电力科学与工程， 2020， 36（8）： 51-56.
	Yao Wanye， Feng Taoming. Transformer detection based on improved YOLOv3［J］. Electric Power Science and Engineering， 2020， 36（8）： 51-56.
[14]	施胜丹，黄金军，朱霄珣，等. 基于声纹SDP-CNN的变压器局部放电模式识别［J］. 电力信息与通信技术， 2022， 20（10）： 105-112.
	Shi Shengdan， Huang Jinjun， Zhu Xiaoxun， et al. Partial discharge pattern recognition on transformer based on voiceprint SDP-CNN［J］. Electric Power Information and Communication Technology， 2022， 20（10）： 105-112.
[15]	崔佳嘉，马宏忠. 基于改进MFCC和3D-CNN的变压器铁心松动故障声纹识别模型［J］. 电机与控制学报， 2022， 26（12）： 150-160.
	Cui Jiajia， Ma Hongzhong. Voiceprint recognition model of transformer core looseness fault based on improved MFCC and 3D-CNN［J］. Electric Machines and Control， 2022， 26（12）： 150-160.
[16]	于达，张玮，王辉. 基于LSTM神经网络的油浸式变压器异常声纹诊断方法研究［J］. 智慧电力， 2023， 51（2）： 45-52.
	Yu Da， Zhang Wei， Wang Hui. Abnormal voiceprint diagnosis method of oil-immersed transformer based on LSTM neural network［J］. Smart Power， 2023， 51（2）： 45-52.
[17]	Chen L， Huang K， Wang W， et al. State evaluation of 500 kV transformer in DC bias based on information fusion［C］// EI2 2022. Piscataway： IEEE， 2022： 1266-1271.
[18]	Henriquez P， Alonso J B， Ferrer M A， et al. Review of automatic fault diagnosis systems using audio and vibration signals［J］. IEEE Transactions on Systems， Man， and Cybernetics： Systems， 2014， 44（5）： 642-652.
[19]	齐子豪，仝杰，张中浩，等. 基于多粒度知识特征和Transformer网络的电力变压器故障声纹辨识方法［J］. 中国电机工程学报， 2025， 45（4）： 1311-1322.
	Qi Zihao， Tong Jie， Zhang Zhonghao， et al. A voiceprint classification method for power transformer fault identification based on multi-granularity knowledge features and Transformer network［J］. Proceedings of the CSEE， 2025， 45（4）： 1311-1322.
[20]	Torky M， Dahy G， Hassanien A E. Recognizing sounds of Red Palm Weevils （RPW） based on the VGGish model： transfer learning methodology［J］. Computers and Electronics in Agriculture， 2023， 212： No.108079.
[21]	Yu Z， Wei Y， Niu B， et al. Automatic condition monitoring and fault diagnosis system for power transformers based on voiceprint recognition［J］. IEEE Transactions on Instrumentation and Measurement， 2024， 73： No.9600411.

故障类型	严重等级	关键参数指标	样本数
直流偏磁	L	直流电流分量5%额定电流，二次谐波含量上升3 dB	500
	M	直流分量10%额定电流，二次谐波含量上升8 dB	500
	H	直流分量20%额定电流，二次谐波含量上升15 dB	500
夹件松动	L	振动基频偏移±2 Hz，幅值波动≤10%	500
	M	基频偏移±5 Hz，幅值波动20%~30%	500
	H	基频偏移±10 Hz，幅值波动≥50%	500
冷却器异响	L	异响频率500 Hz，声压级65 dB（A）	500
	M	异响频率800 Hz，声压级75 dB（A）	500
	H	异响频率1.2 kHz，声压级85 dB（A）	500
短路冲击	L	短路电流2倍额定，持续100 ms	500
	M	短路电流5倍额定，持续300 ms	500
	H	短路电流10倍额定，持续500 ms	500
局部放电	L	局放量50 pC，放电重复率每秒5次	500
	M	局放量200 pC，放电重复率每秒20次	500
	H	局放量800 pC，放电重复率每秒50次	500
重过载	L	负载率130%，温升40 K	500
	M	负载率150%，温升60 K	500
	H	负载率180%，温升80 K	500

故障类型	严重等级	关键参数指标	样本数
直流偏磁	L	直流电流分量5%额定电流，二次谐波含量上升3 dB	500
	M	直流分量10%额定电流，二次谐波含量上升8 dB	500
	H	直流分量20%额定电流，二次谐波含量上升15 dB	500
夹件松动	L	振动基频偏移±2 Hz，幅值波动≤10%	500
	M	基频偏移±5 Hz，幅值波动20%~30%	500
	H	基频偏移±10 Hz，幅值波动≥50%	500
冷却器异响	L	异响频率500 Hz，声压级65 dB（A）	500
	M	异响频率800 Hz，声压级75 dB（A）	500
	H	异响频率1.2 kHz，声压级85 dB（A）	500
短路冲击	L	短路电流2倍额定，持续100 ms	500
	M	短路电流5倍额定，持续300 ms	500
	H	短路电流10倍额定，持续500 ms	500
局部放电	L	局放量50 pC，放电重复率每秒5次	500
	M	局放量200 pC，放电重复率每秒20次	500
	H	局放量800 pC，放电重复率每秒50次	500
重过载	L	负载率130%，温升40 K	500
	M	负载率150%，温升60 K	500
	H	负载率180%，温升80 K	500

N_MFCC	数据集A准确率/%	数据集B准确率/%	数据集C准确率/%
13	96.77±0.42	95.67±0.44	98.12±0.52
20	98.57±0.17	98.38±0.12	99.17±0.15
40	99.68±0.02	99.88±0.02	99.79±0.01

N_MFCC	数据集A准确率/%	数据集B准确率/%	数据集C准确率/%
13	96.77±0.42	95.67±0.44	98.12±0.52
20	98.57±0.17	98.38±0.12	99.17±0.15
40	99.68±0.02	99.88±0.02	99.79±0.01

测试状态	正确识别信号数	测试信号数	准确率/%
总量	1 197	1 200	99.75
正常状态	998	1 000	99.80
故障状态	199	200	99.50

Fault diagnosis method integrating dual-threshold preliminary screening and edge-level dynamic graph convolution

融合双阈值初筛与边级动态图卷积的故障诊断方法

RichHTML

PDF

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 14

References 21

Related Articles 15

Recommended Articles

Metrics

测试状态	正确识别信号数	测试信号数	准确率/%
总量	1 381	1 400	98.64
单一故障	992	1 000	99.20
混合故障	389	400	97.25

故障类型	声音类型	结果
正常（无故障）	正常信号	1 000/1 000
正常（无故障）	正常信号+噪声	1 000/1 000
直流偏磁故障	故障信号	2 000/2 000
夹件松动故障	故障信号
冷却器异响故障	故障信号
短路冲击故障	故障信号+噪声

方法	准确率/%				每帧耗时/s
方法	正常/故障分类	故障识别	数据集含噪声	数据集含混合故障	正常/故障分类	故障识别	数据集含噪声	数据集含混合故障
LSTM	100.00	93.22	83.36	77.28	0.077	18.123	20.230	21.121
VGGish-KNN	100.00	94.25	92.11	93.22	0.067	10.177	17.088	19.120
HMM	100.00	98.16	97.14	96.54	0.034	10.027	10.042	15.127
TOTA	100.00	97.22	93.26	94.23	0.058	18.220	15.172	20.144
GCN	100.00	81.22	65.33	62.37	0.071	22.844	25.558	27.432
本文方法	100.00	100.00	98.94	98.17	0.021	9.014	10.011	11.232

[1]	Yifan WANG, Hu HAN, Dong LI, Yating FAN, Lin LI. Heterogeneous enhancement and multi-source knowledge fusion for aspect-based sentiment analysis [J]. Journal of Computer Applications, 2026, 46(9): 2752-2760.
[2]	Musheng CHEN, Wenqing FU, Xiaohong QIU, Junhua WU, Qiang WEN. Aspect sentiment triplet extraction based on graph convolutional network and cross-domain data augmentation [J]. Journal of Computer Applications, 2026, 46(8): 2432-2439.
[3]	Huihui LYU, Zhaoman ZHONG, Yu ZHANG, Jidong FAN. Syntactically enhanced aspect-based sentiment analysis via mutual information maximization and contrastive learning [J]. Journal of Computer Applications, 2026, 46(8): 2440-2446.
[4]	Pengyu CHEN, Baojun TIAN, Lichang ZHAO, Jiandong FANG. Learning path recommendation model integrating multi-behavior modeling and reinforcement learning [J]. Journal of Computer Applications, 2026, 46(8): 2485-2493.
[5]	Xuechao LIAO, Rui CHEN. Prediction-evaluation framework for anomaly detection in electric vehicle lithium-ion battery [J]. Journal of Computer Applications, 2026, 46(5): 1614-1623.
[6]	Baoyuan ZHENG, Chaobo HE. Graph convolutional network enhanced by graph diffusion and dual-view feature learning [J]. Journal of Computer Applications, 2026, 46(5): 1370-1377.
[7]	Huijie GUO, Tianfeng DOU, Zhenlin ZHANG, Kaiyuan QI, Dong WU, Zhijian QU, Zhao LI, Chongguang REN. Time-interdependency-aware dynamic Bayesian network for traffic prediction [J]. Journal of Computer Applications, 2026, 46(5): 1507-1517.
[8]	Wenhao LI, Yinzhang GUO. Urban traffic flow prediction based on dual-layer multi-scale dynamic graph convolutional network model [J]. Journal of Computer Applications, 2026, 46(4): 1323-1333.
[9]	Xia FENG, Qi WANG, Mingze ZHANG, Haichao ZUO, Guoyu WANG. Flight ground service milestone event prediction based on CTrans-ATGCN [J]. Journal of Computer Applications, 2026, 46(10): 3161-3170.
[10]	Chao SHI, Yuxin ZHOU, Qian FU, Wanyu TANG, Ling HE, Yuanyuan LI. Action recognition algorithm for ADHD patients using skeleton and 3D heatmap [J]. Journal of Computer Applications, 2025, 45(9): 3036-3044.
[11]	Ziliang LI, Guangli ZHU, Yulei ZHANG, Jiajia LIU, Yixuan JIAO, Shunxiang ZHANG. Aspect-based sentiment analysis model integrating syntax and sentiment knowledge [J]. Journal of Computer Applications, 2025, 45(6): 1724-1731.
[12]	Quan WANG, Qixiang LU, Pei SHI. Multi-graph diffusion attention network for traffic flow prediction [J]. Journal of Computer Applications, 2025, 45(5): 1472-1479.
[13]	Man CHEN, Xiaojun YANG, Huimin YANG. Pedestrian trajectory prediction based on graph convolutional network and endpoint induction [J]. Journal of Computer Applications, 2025, 45(5): 1480-1487.
[14]	Yufei LONG, Yuchen MOU, Ye LIU. Multi-source data representation learning model based on tensorized graph convolutional network and contrastive learning [J]. Journal of Computer Applications, 2025, 45(5): 1372-1378.
[15]	Weichao DANG, Chujun SONG, Gaimei GAO, Chunxia LIU. Multi-behavior recommendation based on cascading residual graph convolutional network [J]. Journal of Computer Applications, 2025, 45(4): 1223-1231.