基于线性判别分析和分步机器学习的变压器故障诊断

doi:10.3785/j.issn.1008-973X.2020.11.022

浙江大学学报(工学版)

2020, Vol. 54

Issue (11): 2266-2272 DOI: 10.3785/j.issn.1008-973X.2020.11.022

电气工程

基于线性判别分析和分步机器学习的变压器故障诊断

谢乐(

),衡熙丹,刘洋,蒋启龙,刘东*(

)

西南交通大学电气工程学院，四川成都 611756

Transformer fault diagnosis based on linear discriminant analysis and step-by-step machine learning

Le XIE(

),Xi-dan HENG,Yang LIU,Qi-long JIANG,Dong LIU*(

)

School of Electrical Engineering, Southwest Jiaotong University, Chengdu 611756, China

全文: PDF(1346 KB) HTML

摘要：

为了改善当前变压器故障诊断在特征量选取和使用单一诊断模型进行故障诊断上的不足，提高变压器故障诊断的准确率和效率，提出基于线性判别分析（LDA）的特征选取方法，建立基于分步机器学习的诊断模型. 该模型选取16组油中溶解气体体积分数比值的多特征参数，运用线性判别分析对参数进行降维作为输入特征向量；运用概率神经网络对变压器故障做出初步诊断，区分出易混淆故障；使用基于灰狼群算法优化的支持向量机对易混淆故障做进一步的区分. 最终实验诊断准确率为97.27%，诊断时间为4.87 s. 与单一机器学习模型相比，所提出的模型不仅具有更高的准确率，还具有更高的效率. 实例分析表明，本研究方法能有效弥补单一机器学习的缺陷，为故障样本有限情况下的电力变压器故障诊断提供参考.

关键词： 变压器; 故障诊断; 机器学习; 特征参数; 线性判别分析(LDA)

Abstract:

A feature selection method based on the linear discriminant analysis (LDA) was proposed, and a diagnostic model based on the step-by-step machine learning was established, in order to improve the current shortcomings of transformer fault diagnosis in feature selection and single diagnosis model, as well as the accuracy and the efficiency of transformer fault diagnosis. The multi-characteristic parameters of volume fraction ratio of dissolved gas in 16 groups of oil were selected by this model. And the linear discriminant analysis was performed to apply dimensionality reduction on parameters and the results were used as input eigenvectors. Then the probabilistic neural networks were used to diagnose transformer faults and to distinguish confusing faults. The confusing fault was further distinguished by the support vector machine optimized by the grey wolf swarm algorithm.The final experimental diagnostic accuracy rate was 97.27%, while the diagnostic time was 4.87 s. The proposed model not only has higher accuracy, but also has better efficiency, compared with a single machine learning model. Case analysis shows that this method can make up for the shortcoming of single machine learning, which can provide reference for the fault diagnosis of power transformer with limited fault cases.

Key words: transformer fault diagnosis machine learning characteristic parameter linear discriminant analysis (LDA)

收稿日期: 2019-11-20 出版日期: 2020-12-15

CLC:

TM 411

基金资助: 国家自然科学基金资助项目（61531016）；国家自然科学联合基金资助项目（U1934204）

通讯作者: 刘东 E-mail: leohfut@126.com;liudong@swjtu.edu.cn

作者简介: 谢乐（1997—），男，硕士生，从事智能算法与故障诊断研究. orcid.org/0000-0002-5357-648X. E-mail： leohfut@126.com

	服务
	把本文推荐给朋友
	加入引用管理器
	E-mail Alert
	作者相关文章
	谢乐
	衡熙丹
	刘洋
	蒋启龙
	刘东

引用本文:

谢乐,衡熙丹,刘洋,蒋启龙,刘东. 基于线性判别分析和分步机器学习的变压器故障诊断[J]. 浙江大学学报(工学版), 2020, 54(11): 2266-2272.

Le XIE,Xi-dan HENG,Yang LIU,Qi-long JIANG,Dong LIU. Transformer fault diagnosis based on linear discriminant analysis and step-by-step machine learning. Journal of ZheJiang University (Engineering Science), 2020, 54(11): 2266-2272.

链接本文:

http://www.zjujournals.com/eng/CN/10.3785/j.issn.1008-973X.2020.11.022 或 http://www.zjujournals.com/eng/CN/Y2020/V54/I11/2266

表 1 多特征参数

图 1 灰狼群优化支持向量机流程图

图 2 分步机器学习诊断模型

表 2 故障状态标记

图 3 原始数据的分维可视化图

图 4 原始数据降维后的分维可视化图

图 5 线性判别分析的降维效果

图 6 概率神经网络诊断结果

图 7 灰狼群优化支持向量机分类结果

图 8 灰狼群适应度曲线

表 3 本研究方法与其他故障诊断模型的正确率和时间

表 4 变压器故障诊断实例

1	汪永华. 常用电气与电控设备故障诊断400例 [M]. 北京: 中国电力出版社, 2011.
2	ELDIN A A H, REFAEY M A A novel algorithm for discrimination between inrush current and internal faults in power transformer differential protection based on discrete wavelet transform[J]. Electric Power Systems Research, 2011, 81 (1): 19- 24 doi: 10.1016/j.jpgr.2010.07.010
3	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]. 电子测量与仪器学报, 2017, 31 (1): 118- 124 XIA Fei, LUO Zhi-jiang, ZHANG Hao, et al Application of hybrid neural network in transformer fault diagnosis[J]. Journal of Electronic Measurement and Instrument, 2017, 31 (1): 118- 124
4	公茂法, 张言攀, 柳岩妮, 等基于 BP 网络算法优化模糊Petri 网的电力变压器故障诊断[J]. 电力系统保护与控制, 2015, 43 (3): 113- 117 GONG Mao-fa, ZHANG Yan-pan, LIU Yan-ni, et al Power transformer fault diagnosis based on BP neural network optimized fuzzy petri net[J]. Power System Protection and Control, 2015, 43 (3): 113- 117 doi: 10.7667/j.issn.1674-3415.2015.03.017
5	薛浩然, 张珂珩, 李斌, 等基于布谷鸟算法和支持向量机的变压器故障诊断[J]. 电力系统保护与控制, 2015, 43 (8): 8- 13 XUE Hao-ran, ZHANG Ke-heng, LI Bin, et al Transformer fault diagnosis based on cuckoo algorithm and support vector machine[J]. Power System Protection and Control, 2015, 43 (8): 8- 13 doi: 10.7667/j.issn.1674-3415.2015.08.002
6	张耀鹏. 基于决策树算法的扼流适配变压器故障诊断系统研究与设计[D]. 北京: 北京交通大学, 2018. ZHANG Yao-peng. Research and design of fault diagnosis system for turbulence adaptive transformer based on decision tree algorithm [D]. Beijing: Beijing Jiaotong University, 2018.
7	吐松江·卡日, 高文胜, 张紫薇, 等基于支持向量机和遗传算法的变压器故障诊断[J]. 清华大学学报: 自然科学版, 2018, 58 (7): 623- 629 TU Song-jiang·KA Ri, GAO Wen-sheng, ZHANG Zi-wei, et al Transformer fault diagnosis based on support vector machine and genetic algorithm[J]. Journal of Tsinghua University: Science and Technology, 2018, 58 (7): 623- 629
8	汪可, 李金忠, 张书琦, 等变压器故障诊断用油中溶解气体新特征参量[J]. 中国电机工程学报, 2016, 36 (23): 6570- 6578 WANG Ke, LI Jin-zhong, ZHANG Shu-qi, et al New characteristic parameters of dissolved gases in oil for transformer fault diagnosis[J]. Proceedings of the CSEE, 2016, 36 (23): 6570- 6578
9	高杰. 基于优化PNN网络的变压器故障诊断研究[D]. 淮南: 安徽理工大学, 2018. GAO Jie. Research on transformer fault diagnosis based on optimized PNN network [D]. Huainan: Anhui University of Science and Technology, 2018.
10	YANG X, CHEN W, LI A, et al BA-PNN-based methods for power transformer fault diagnosis[J]. Advanced Engineering Informatics, 2019, 39: 178- 185 doi: 10.1016/j.aei.2019.01.001
11	吴广宁, 袁海满, 宋臻杰, 等基于粗糙集与多类支持向量机的电力变压器故障诊断[J]. 高电压技术, 2017, 43 (11): 3668- 3674 WU Guang-ning, YUAN Hai-man, SONG Zhen-jie, et al Fault diagnosis of power transformer based on rough set and multi-class support vector machine[J]. High Voltage Engineering, 2017, 43 (11): 3668- 3674
12	阳艳电力变压器常见故障分析与检测[J]. 中国科技博览, 2015, (26): 278 YANG Yan Analysis and detection of common faults in power transformers[J]. China Science and Technology Expo, 2015, (26): 278
13	DUVAL M, DEPABLA A Interpretation of gas-in-oil analysis using new IEC publication 60599 and IEC TC 10 databases[J]. IEEE Electrical Insulation Magazine, 2001, 17 (2): 31- 41 doi: 10.1109/57.917529
14	BAKAR N A, ABU-SIADA A, ISLAM S A review of dissolved gas analysis measurement and interpretation techniques[J]. IEEE Electrical Insulation Magazine, 2014, 30 (3): 39- 49 doi: 10.1109/MEI.2014.6804740
15	ROGERS R R IEEE and IEC codes to interpret incipient faults in transformers, using gas in oil analysis[J]. IEEE Transactions on Electrical Insulation, 1978, (5): 349- 354
16	王学磊, 李庆民, 杨芮, 等基于油色谱分析的变压器复合绝缘缺陷多指标综合权重评估方法[J]. 高电压技术, 2015, 41 (11): 3836- 3842 WANG Xue-lei, LI Qing-min, YANG Rui, et al Multi-index comprehensive weight evaluation method for transformer composite insulation defects based on oil chromatography[J]. High Voltage Engineering, 2015, 41 (11): 3836- 3842
17	李春茂, 周妺末, 刘亚婕, 等基于邻域粗糙集与多核支持向量机的变压器多级故障诊断[J]. 高电压技术, 2018, 44 (11): 3474- 3482 LI Chun-mao, ZHOU Mo-mo, LIU Ya-jie, et al Multi-level fault diagnosis of transformer based on neighborhood rough set and multi-core support vector machine[J]. High Voltage Technology, 2018, 44 (11): 3474- 3482
18	周志华. 机器学习 [M]. 北京: 清华大学出版社, 2016: 230-232.
19	LI M, YUAN B 2D-LDA: astatistical linear discriminant analysis for image matrix[J]. Pattern Recognition Letters, 2005, 26 (5): 527- 532 doi: 10.1016/j.patrec.2004.09.007
20	SPECHT D F Probabilistic neural networks[J]. Neural Networks, 1990, 3 (1): 109- 118 doi: 10.1016/0893-6080(90)90049-Q
21	HEARST M A, DUMAIS S T, OSUUA E, et al Support vector machines[J]. IEEE Intelligent Systems and Their Applications, 1998, 13 (4): 18- 28 doi: 10.1109/5254.708428
22	魏政磊, 赵辉, 李牧东, 等控制参数值非线性调整策略的灰狼优化算法[J]. 空军工程大学学报: 自然科学版, 2016, 17 (3): 104- 110 WEI Zheng-lei, ZHAO Hui, LI Mu-dong, et al Grey wolf optimization algorithm for controlling parameter value nonlinear adjustment strategy[J]. Journal of Air Force Engineering University, 2016, 17 (3): 104- 110
23	IEC. Mineral oil-impregnated electrical equipment in service-guide to the interpretation of dissolved and free gases analysis: EN 60599-2007 [S]. Geneva: IEC/CEI, 2007.

[1]	战友,李强,马啸天,王郴平,邱延峻. 基于宏微观纹理特征融合的路面摩擦性能预测[J]. 浙江大学学报(工学版), 2021, 55(4): 684-694.
[2]	于勇,薛静远,戴晟,鲍强伟,赵罡. 机加零件质量预测与工艺参数优化方法[J]. 浙江大学学报(工学版), 2021, 55(3): 441-447.
[3]	陈巧红,陈翊,李文书,贾宇波. 多尺度SE-Xception服装图像分类[J]. 浙江大学学报(工学版), 2020, 54(9): 1727-1735.
[4]	郭燕妮,杨遥,黄正梁,王靖岱,阳永荣. 缠绕管结构参数对管内压降的影响及压降模型[J]. 浙江大学学报(工学版), 2020, 54(8): 1620-1627.
[5]	金列俊,詹建明,陈俊华,王涛. 基于一维卷积神经网络的钻杆故障诊断[J]. 浙江大学学报(工学版), 2020, 54(3): 467-474.
[6]	王慧芳,张晨宇. 采用极限梯度提升算法的电力系统电压稳定裕度预测[J]. 浙江大学学报(工学版), 2020, 54(3): 606-613.
[7]	林京京,沈艳霞. 永磁同步电机驱动系统电流传感器容错控制[J]. 浙江大学学报(工学版), 2019, 53(9): 1815-1825.
[8]	张政,周学军,王希晨,周媛媛. 缆系水下信息网恒流远供系统短路故障诊断及区间定位方法[J]. 浙江大学学报(工学版), 2019, 53(6): 1190-1197.
[9]	万志远,陶嘉恒,梁家坤,才振功,苌程,乔林,周巧妮. Stack Overflow上机器学习相关问题的大规模实证研究[J]. 浙江大学学报(工学版), 2019, 53(5): 819-828.
[10]	柯懂湘,潘丽敏,罗森林,张寒青. 基于随机森林算法的Android恶意行为识别与分类方法[J]. 浙江大学学报(工学版), 2019, 53(10): 2013-2023.
[11]	忽丽莎, 王素贞, 陈益强, 高晨龙, 胡春雨, 蒋鑫龙, 陈振宇, 高兴宇. 基于可穿戴设备的跌倒检测算法综述[J]. 浙江大学学报(工学版), 2018, 52(9): 1717-1728.
[12]	苏国东, 孙玲玲, 王翔, 王尊峰, 张胜洲, 雷宇超. 126.6~128.1 GHz基波压控振荡器设计[J]. 浙江大学学报(工学版), 2018, 52(9): 1788-1795.
[13]	童水光, 张依东, 徐剑, 从飞云. 频带多尺度复合模糊熵及其在轴承故障诊断中的应用[J]. 浙江大学学报(工学版), 2018, 52(8): 1509-1516.
[14]	葛云龙, 陈自强, 郑昌文. UTSTF锂离子电池时变参数估计与故障诊断[J]. 浙江大学学报(工学版), 2018, 52(6): 1223-1230.
[15]	王洪凯, 陈中华, 周纵苇, 李迎辞, 陆佩欧, 王文志, 刘宛予, 于丽娟. 机器学习算法诊断PET/CT纵膈淋巴结性能评估[J]. 浙江大学学报(工学版), 2018, 52(4): 788-797.

Viewed

Full text

Abstract

Cited

Shared

Discussed