Research on fault diagnosis method of power transformers based on multi-scale extreme fusion network

doi:10.3969/j.issn.2097-0706.2022.09.011

Abstract

Abstract:

To cope with the low accuracy of the power transformer fault diagnosis taking a classification method with fixed-parameter algorithm,a new diagnosis method based on multi-scale extreme fusion network（MEFN）is proposed.There are two main steps in this method.Firstly,the initial diagnosis results are generated by extreme learning machine （ELM） models with different parameter scales.Models are made to analyze the same set of data about dissolved gas,and reflect different statistical characteristics,providing complementary statistical information to power transformer fault diagnosis.Then,the diagnostic results are fused by the decision-level information fusion algorithm,which can improve the accuracy of the power transformer fault diagnosis.Modeling training and test for the proposed method are made based on 487 sets of experimental data.The test results show that,MEFN can detect six types of common transformer faults with a high recognition accuracy of 94%.The recognition accuracy of MEFN is higher than that of support vector machine （SVM） and back-propagation （BP） neural network by 8% and 13%,respectively.The proposed method can meet the requirements on transformer fault diagnosis made by power companies.

Key words: power transformer, fault diagnosis,smart grid, ELM, information fusion, MEFN

CLC Number:

TK01：TM933

LONG Sicheng, HUANG Zhihong. Research on fault diagnosis method of power transformers based on multi-scale extreme fusion network[J]. Integrated Intelligent Energy, 2022, 44(9): 78-83.

Figures/Tables 8

Fig.1

Table 1

Table 2

Fig.2

Table 3

Table 4

Table 5

Fig.3

References 20

[1]	魏佳栋, 许飞, 曹辉, 等. 基于物联网的配电变压器智能感知平台及其安全构架[J]. 华电技术, 2021, 43(1):1-5.
	WEI Jiadong, XU Fei, CAO Hui, et al. Intelligent perception platform for distribution transformers and its safety architecture based on IoT[J]. Huadian Technology, 2021, 43(1):1-5.
[2]	郑含博, 王伟, 李晓纲, 等. 基于多分类最小二乘支持向量机和改进粒子群优化算法的电力变压器故障诊断方法[J]. 高电压技术, 2014, 40(11):3424-3429.
	ZHENG Hanbo, WANG Wei, LI Xiaogang, et al. Fault diagnosis method of power transformers using multi‑class LS‑SVM and improved PSO[J]. High Voltage Engineering, 2014, 40(11):3424-3429.
[3]	师瑞峰, 史永锋, 牟军, 等. 油中溶解气体电力变压器故障诊断专家系统[J]. 电力系统及其自动化学报, 2014, 26(12):49-54.
	SHI Ruifeng, SHI Yongfeng, MU Jun, et al. Power transformer fault diagnosis expert system with dissolved gas analysis in oil[J]. Proceedings of the CSU-EPSA, 2014, 26(12):49-54.
[4]	张镱议, 彭鸿博, 李昕, 等. 基于DGA特征量优选与改进磷虾群算法优化支持向量机的变压器故障诊断模型[J]. 电测与仪表, 2019, 56(21):110-116.
	ZHANG Yiyi, PENG Hongbo, LI Xin, et al. Fault diagnosis model of power transformers based on DGA features selection and improved krill herd algorithm optimization SVM[J]. Electrical Measurement & Instrumentation, 2019, 56(21):110-116.
[5]	丁硕, 常晓恒, 巫庆辉, 等. DGA与GRNN的联合变压器故障诊断研究[J]. 电子测量技术, 2014, 37(5):142-146.
	DING Shuo, CHANG Xiaoheng, WU Qinhui, et al. Study of transformer fault diagnosis based on GRNN and DGA[J]. Electronic Measurement Technology, 2014, 37(5):142-146.
[6]	DORNERBURG E, STRITTMATTTER W. Monitoring oil‑cooled transformers by gas‑analysis[J]. Brown Boveri Review, 1974, 61(5):238-247.
[7]	DUVAL M. New techniques for dissolved gas‑in‑oil analysis[J]. IEEE Electrical Insulation Magazine, 2003, 19(2):6-15.
[8]	WU H, LI X, WU D. RMP neural network based dissolved gas analyzer for fault diagnostic of oil‑filled electrical equipment[J]. IEEE Transactions on Dielectrics and Electrical Insulation, 2011, 18(2):495-498. doi: 10.1109/TDEI.2011.5739454
[9]	贾亦敏, 史丽萍, 严鑫. 基于精英混沌蜂群算法优化小波神经网络的变压器故障诊断[J]. 高压电器, 2020, 56(8):230-236.
	JIA Yimin, SHI Liping, YAN Xin. Transformer fault diagnosis using wavelet neural network based on elite‑chaos artificial bee colony algorithm[J]. High Voltage Apparatus, 2020, 56(8):230-236.
[10]	公茂法, 柳岩妮, 王来河, 等. 基于混沌优化粒子群BP神经网络的电力变压器故障诊断[J]. 电测与仪表, 2016, 53(15):13-16,32.
	GONG Maofa, LIU Yanni, WANG laihe, et al. Fault diagnosis of power transformers based on chaos particle swarm optimization BP neural network[J]. Electrical Measurement & Instrumentation, 2016, 53(15):13-16,32.
[11]	夏飞, 罗志疆, 张浩, 等. 混合神经网络在变压器故障诊断中的应用[J]. 电子测量与仪器, 2017, 31(1):118-124.
	XIA Fei, LUO Zhijiang, ZHANG Hao, et al. Application of mixed neural network in transformer fault diagnosis[J]. Journal of Electronic Measurement and Instrumentation, 2017, 31(1):118-124.
[12]	周锋, 孙廷玺, 权少静, 等. 基于集合经验模态分解和极限学习机的变压器油中溶解气体体积分数预测方法[J]. 高电压技术, 2020, 46(10):3658-3665.
	ZHOU Feng, SUN Yanxi, QUAN Shaojin, et al. Predication of dissolved gases concentration in transformer oil based on ensemble empirical mode decomposition and extreme learning machine[J]. High Voltage Engineering, 2020, 46(10):3658-3665.
[13]	王春明, 朱永利. 基于深度降噪极限学习机的变压器故障诊断[J]. 电测与仪表, 2019, 56(15):143-147.
	WANG Chunming, ZHU Yongli. Transformer fault diagnosis based on deep de‑noising extreme learning machine[J]. Electrical Measurement & Instrumentation, 2019, 56(15):143-147.
[14]	雷帆, 高波, 袁海满, 等. 基于DGA的粗糙集与人工鱼群极限学习机的变压器故障诊断[J]. 高压电器, 2017, 53(10):124-130.
	LEI Fan, GAO Bo, YUAN Haiman, et al. Transformer fault diagnosis by using rough set and artificial fish swarm extreme learning machine based on DGA[J]. High Voltage Apparatus, 2017, 53(10):124-130.
[15]	LI S, KANG X, HU J. Image fusion with guided filtering[J]. IEEE Transactions on Image processing, 2013, 22(7):2864-2875. doi: 10.1109/TIP.2013.2244222
[16]	DUVAL M. A review of faults detectable by gas‑in‑oil analysis in transformers[J]. IEEE Electrical Insulation magazine, 2002, 18(3):8-17.
[17]	黄志鸿, 毛建旭, 王耀南, 等. 基于机器视觉的啤酒瓶口缺陷检测分类方法研究[J]. 电子测量与仪器学报, 2016, 30(6):873-879.
	HUANG Zhihong, MAO Jianxu, WANG Yaonan, et al. Research on beer bottle defect classification detection method based on machine vision[J]. Journal of Electronic Measurement and Instrumentation, 2016, 30(6):873-879.
[18]	蒋昀宸, 樊绍胜, 陈骏星溆. 带电作业智能新技术及其应用现状[J]. 湖南电力, 2018, 38(5):1-4.
	JIANG Yunchen, FAN Zhaosheng, CHEN Junxingxu, ei al. Smart new‑technologies and applications for live work[J]. Hunan Electric Power, 2018, 38(5):1-4.
[19]	贾志军, 白德龙, 宋燕杰, 等. 基于KPCA算法的电厂设备在线故障诊断研究[J]. 华电技术, 2021, 43(8):48-53.
	JIA Zhijun, BAI Delong, SONG Yanjie, et al. Research on on‑line fault diagnosis and treatment of power plant equipment based on KPCA[J]. Huadian Technology, 2021, 43(8):48-53.
[20]	胡杰, 唐静, 谢仕义, 等. 基于大数据技术的电厂设备状态评估和预警应用研究[J]. 华电技术, 2020, 42(2):1-6.
	HU Jie, TANG Jing, XIE Shiyi, et al. State assessment and early warning application for power plant equipment based on big data technology[J]. Huadian Technology, 2020, 42(2):1-6.

气体体积分数	编码
气体体积分数	C₂H₂/C₂H₄	CH₄/H₂	C₂H₄/C₂H₆
<0.1	0	1	0
≥0.1~1.0	1	0	0
≥1.0~3.0	1	2	1
≥3.0	2	2	2

故障类型	气体编码
故障类型	C₂H₂/C₂H₄	CH₄/H₂	C₂H₄/C₂H₆
正常	0	0	0
局部放电	0或1	1	0
低压放电	1或2	0	0或1或2
高压放电	1	0	2
低温过载	0	0	1
中温过载	0	2	0
高温过载	0	2	1或2

样本序号	模型						真实故障类型
样本序号	ELM₅₀	ELM₁₀₀	ELM₁₅₀	ELM₂₀₀	ELM₂₅₀	MEFN	真实故障类型
1	PD	D₁	D₁	D₁	D₁	D₁	D₁
2	D₂	D₂	D₂	D₃	D₂	D₂	D₂
3	D₁	D₂	D₃	D₃	D₃	D₃	D₃
4	T₁	T₂	T₁	T₁	T₁	T₁	T₁
5	T₂	T₃	T₃	T₃	T₃	T₃	T₃
6	PD	PD	D₂	PD	PD	PD	PD
7	D₁	D₁	D₁	D₁	D₂	D₁	D₁

故障类型	BP分类器			SVM分类器			MEFN分类器
故障类型	N_total	N_right	准确率/%	N_total	N_right	准确率/%	N_total	N_right	准确率/%
PD	15	12	80.00	15	11	73.33	15	14	93.33
D₁	13	11	84.62	13	12	92.31	13	13	100.00
D₂	6	3	50.00	6	5	83.33	6	5	83.33
T₁	19	14	73.68	19	15	78.95	19	17	89.47
T₂	23	22	95.65	23	20	86.96	23	22	95.65
T₃	8	8	62.50	8	7	87.50	8	8	100.00
正常	16	13	81.25	16	15	93.75	16	15	93.75