考虑样本类内不平衡的CHPOA-DBN变压器故障诊断方法

doi:10.11930/j.issn.1004-9649.202305039

中国电力 ›› 2023, Vol. 56 ›› Issue (10): 133-144.DOI: 10.11930/j.issn.1004-9649.202305039

考虑样本类内不平衡的CHPOA-DBN变压器故障诊断方法

王爽¹^,²(), 罗倩¹^,²(), 唐波¹^,²(), 姜岚¹^,²(), 李锦³()

1. 三峡大学电气与新能源学院，湖北宜昌　443002
2. 三峡大学湖北省输电线路工程技术研究中心，湖北宜昌　443002
3. 国网湖北省电力有限公司超高压公司，湖北武汉　430051

收稿日期:2023-05-09 出版日期:2023-10-28 发布日期:2023-10-31
作者简介:王爽（1987—），男，通信作者，博士，讲师，从事电力设备故障诊断与状态评估研究，E-mail： wsctp168@126.com
罗倩（1998—），女，硕士研究生，从事电力设备故障诊断与状态评估研究，E-mail： luoqian19982021@163.com
唐波（1978—），男，博士，教授，从事输变电工程电磁环境方向研究，E-mail： tangboemail@sina.com
姜岚（1986—），男，博士，讲师，从事输电线路工程结构振动与控制研究，E-mail： jl@ctgu.edu.cn
李锦（1985—），男，硕士，工程师，从事变电运维技术研究，E-mail： 573239771@qq.com
基金资助:
湖北省重点研发计划资助项目（2020BAB110）；电力系统及大型发电设备安全控制与仿真国家重点实验室开放基金项目（SKLD21KM11）。

CHPOA-DBN Transformer Fault Diagnosis Method Considering Sample Within-Class Imbalance

Shuang WANG¹^,²(), Qian LUO¹^,²(), Bo TANG¹^,²(), Lan JIANG¹^,²(), Jin LI³()

1. School of Electrical Engineering and New Energy, China Three Gorges University, Yichang 443002, China
2. Hubei Provincial Engineering Technology Research Center for Power Transmission Line, China Three Gorges University, Yichang 443002, China
3. State Grid Hubei Extra High Voltage Company, Wuhan 430051, China

Received:2023-05-09 Online:2023-10-28 Published:2023-10-31
Supported by:
This work is supported by Key Research and Development Projects of Hubei Province (No.2020BAB110) and Open Foundation Project of State Key Laboratory for Safety Control and Simulation of Power System and Large Power Generation Equipment (No.SKLD21KM11).

摘要/Abstract

摘要：

为解决变压器故障样本类内不平衡与人为确定深度信念网络（deep belief network，DBN）的网络参数导致故障诊断精度低的问题，提出一种基于样本均衡和改进DBN的变压器故障诊断方法。首先，针对合成少数类过采样算法（synthesis minority oversampling technique，SMOTE）生成样本加剧类内不平衡的问题，提出基于改进K均值（improved K-means，IK-means）的IK-means SMOTE算法，据此得到类间、类内均衡的故障样本；其次，利用Tent混沌映射改进的鹈鹕优化算法（chaotic hybrid pelican optimization algorithm，CHPOA）对DBN的隐含层节点数、反向微调学习率寻优，构建CHPOA-DBN变压器故障诊断模型；最后，基于实验数据，分别将经典过采样算法、经典故障诊断模型与所提方法进行对比分析，结果表明：所提方法故障诊断准确率达到96.25%，可以为变压器故障样本不均衡条件下的故障智能诊断提供重要参考。

关键词: 变压器故障诊断, 类内不平衡, 样本均衡, Tent混沌映射, DBN网络参数寻优

Abstract:

In recent years, deep belief network (DBN) based transformer fault diagnosis methods have been developed. However, they share two prominent drawbacks, which are the low accuracy issue caused by the within-class imbalance of transformer faults samples and the artificial determination of the network parameters of deep belief network (DBN). In this paper, a transformer fault diagnosis method based on sample balance processing and improved DBN is proposed. Firstly, an improved K-means (IK-means) synthesis minority oversampling technique (SMOTE) algorithm is proposed to obtain within-class and between-class balanced fault samples. Then, the Tent chaotic map embedded chaotic hybrid pelican optimization algorithm (CHPOA) is developed to optimize the number of hidden layer nodes and reverse fine-tuning learning rate of DBN, and the CHPOA-DBN transformer fault diagnosis model is constructed. Finally, the classical oversampling algorithm, the classical fault diagnosis model and the proposed method are compared and analyzed, based on the experimental data, respectively. The results show that the fault diagnosis accuracy of the proposed method reaches 96.25 %, which provide an important reference for intelligent fault diagnosis under imbalanced fault samples of transformers.

Key words: transformer fault diagnosis, within-class imbalance, sample balance processing, Tent chaotic map, DBN network parameters optimization

王爽, 罗倩, 唐波, 姜岚, 李锦. 考虑样本类内不平衡的CHPOA-DBN变压器故障诊断方法[J]. 中国电力, 2023, 56(10): 133-144.

Shuang WANG, Qian LUO, Bo TANG, Lan JIANG, Jin LI. CHPOA-DBN Transformer Fault Diagnosis Method Considering Sample Within-Class Imbalance[J]. Electric Power, 2023, 56(10): 133-144.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: https://www.electricpower.com.cn/CN/10.11930/j.issn.1004-9649.202305039

https://www.electricpower.com.cn/CN/Y2023/V56/I10/133

图/表 14

参考文献 42

1	谢毓城. 电力变压器手册[M]. 北京: 机械工业出版社, 2014: 2–3.
2	LI J Z, ZHANG Q G, WANG K, et al. Optimal dissolved gas ratios selected by genetic algorithm for power transformer fault diagnosis based on support vector machine[J]. IEEE Transactions on Dielectrics and Electrical Insulation, 2016, 23 (2): 1198- 1206. DOI
3	郝玲玲, 朱永利, 王永正. 基于DCAE-KSSELM的变压器故障诊断方法[J]. 中国电力, 2022, 55 (2): 125- 130.
	HAO Lingling, ZHU Yongli, WANG Yongzheng. Transformer fault diagnosis method based on DCAE-KSSELM[J]. Electric Power, 2022, 55 (2): 125- 130.
4	郑业爽, 李世春, 鲁玲. 基于多策略ISOA优化SVM的变压器故障诊断研究[J]. 智慧电力, 2023, 51 (2): 38- 44. DOI
	ZHENG Yeshuang, LI Shichun, LU Ling. Transformer fault diagnosis based on multi-strategy ISOA optimized SVM[J]. Smart Power, 2023, 51 (2): 38- 44. DOI
5	张卫华, 苑津莎, 张铁峰, 等. 应用B样条理论改进的变压器三比值故障诊断方法[J]. 中国电机工程学报, 2014, 34 (24): 4129- 4136. DOI
	ZHANG Weihua, YUAN Jinsha, ZHANG Tiefeng, et al. An improved three-ratio method for transformer fault diagnosis using B-spline theory[J]. Proceedings of the CSEE, 2014, 34 (24): 4129- 4136. DOI
6	TAHA I B M, HOBALLAH A, GHONEIM S S M. Optimal ratio limits of Rogers’ four-ratios and IEC 60599 code methods using particle swarm optimization fuzzy-logic approach[J]. IEEE Transactions on Dielectrics and Electrical Insulation, 2020, 27 (1): 222- 230. DOI
7	IRUNGU G K, AKUMU A O, MUNDA J L. A new fault diagnostic technique in oil-filled electrical equipment; the dual of Duval triangle[J]. IEEE Transactions on Dielectrics and Electrical Insulation, 2016, 23 (6): 3405- 3410. DOI
8	李雅欣, 侯慧娟, 张立静, 等. 近邻成分分析和k近邻学习融合的变压器不平衡样本故障诊断[J]. 高电压技术, 2021, 47 (2): 472- 479. DOI
	LI Yaxin, HOU Huijuan, ZHANG Lijing, et al. Transformer fault diagnosis with unbalanced samples based on neighborhood component analysis and k-nearest neighbors[J]. High Voltage Engineering, 2021, 47 (2): 472- 479. DOI
9	徐龙舞, 张英, 张倩, 等. 基于正交实验法改进的蝠鲼算法优化BP在变压器故障诊断上的研究[J]. 南方电网技术, 2022, 16 (7): 46- 54.
	XU Longwu, ZHANG Ying, ZHANG Qian, et al. Orthogonal experiment method based improved MRFO algorithm to optimize BP in transformer fault diagnosis[J]. Southern Power System Technology, 2022, 16 (7): 46- 54.
10	李云淏, 咸日常, 张海强, 等. 基于改进灰狼算法与最小二乘支持向量机耦合的电力变压器故障诊断方法[J]. 电网技术, 2023, 47 (4): 1470- 1478. DOI
	LI Yunhao, XIAN Richang, ZHANG Haiqiang, et al. Fault diagnosis for power transformers based on improved grey wolf algorithm coupled with least squares support vector machine[J]. Power System Technology, 2023, 47 (4): 1470- 1478. DOI
11	李亮, 范瑾, 闫林, 等. 基于混合采样和支持向量机的变压器故障诊断[J]. 中国电力, 2021, 54 (12): 150- 155.
	LI Liang, FAN Jin, YAN Lin, et al. Transformer fault diagnosis based on hybrid sampling and support vector machines[J]. Electric Power, 2021, 54 (12): 150- 155.
12	刘展程, 王爽, 唐波. 基于布谷鸟搜索算法和DBN模型的变压器故障识别[J]. 电力科学与技术学报, 2022, 37 (2): 3- 11.
	LIU Zhancheng, WANG Shuang, TANG Bo. Transformer fault identification based on the cuckoo search algorithm and DBN model[J]. Journal of Electric Power Science and Technology, 2022, 37 (2): 3- 11.
13	刘云鹏, 许自强, 李刚, 等. 人工智能驱动的数据分析技术在电力变压器状态检修中的应用综述[J]. 高电压技术, 2019, 45 (2): 337- 348. DOI
	LIU Yunpeng, XU Ziqiang, LI Gang, et al. Review on applications of artificial intelligence driven data analysis technology in condition based maintenance of power transformers[J]. High Voltage Engineering, 2019, 45 (2): 337- 348. DOI
14	葛磊蛟, 廖文龙, 王煜森, 等. 数据不足条件下基于改进自动编码器的变压器故障数据增强方法[J]. 电工技术学报, 2021, 36 (S1): 84- 94.
	GE Leijiao, LIAO Wenlong, WANG Yusen, et al. Data augmentation method for transformer fault based on improved auto-encoder under the condition of insufficient data[J]. Transactions of China Electrotechnical Society, 2021, 36 (S1): 84- 94.
15	栗磊, 王廷涛, 赫嘉楠, 等. 考虑过采样器与分类器参数优化的变压器故障诊断策略[J]. 电力自动化设备, 2023, 43 (1): 209- 217. DOI
	LI Lei, WANG Tingtao, HE Jianan, et al. Transformer fault diagnosis strategy considering parameter optimization of oversampler and classifier[J]. Electric Power Automation Equipment, 2023, 43 (1): 209- 217. DOI
16	陈铁, 冷昊伟, 李咸善, 等. 基于油中气体分析与类重叠特征的变压器分层故障诊断模型[J]. 中国电力, 2022, 55 (7): 22- 32, 41.
	CHEN Tie, LENG Haowei, LI Xianshan, et al. Transformer hierarchical fault diagnosis model based on dissolved gas analysis of insulating oil and class overlap features[J]. Electric Power, 2022, 55 (7): 22- 32, 41.
17	LIN W C, TSAI C F, HU Y H, et al. Clustering-based undersampling in class-imbalanced data[J]. Information Sciences, 2017, 409 (26): 17- 26.
18	CHAWLA N V, BOWYER K W, HALL L O, et al. SMOTE: synthetic minority over-sampling technique[J]. Journal of Artificial Intelligence Research, 2002, 16, 321- 357. DOI
19	HAN H, WANG W Y, MAO B H. Borderline-SMOTE: a new over-sampling method in imbalanced data sets learning[C]//Lecture Notes in Computer Science. Berlin, Heidelberg: Springer Berlin Heidelberg, 2005: 878–887.
20	余松, 胡东, 唐超, 等. 基于TLR-ADASYN平衡化数据集的MSSA-SVM变压器故障诊断[J]. 高电压技术, 2021, 47 (11): 3845- 3853. DOI
	YU Song, HU Dong, TANG Chao, et al. MSSA-SVM transformer fault diagnosis method based on TLR-ADASYN balanced data set[J]. High Voltage Engineering, 2021, 47 (11): 3845- 3853. DOI
21	刘云鹏, 和家慧, 许自强, 等. 基于SVM SMOTE的电力变压器故障样本均衡化方法[J]. 高电压技术, 2020, 46 (7): 2522- 2529.
	LIU Yunpeng, HE Jiahui, XU Ziqiang, et al. Equalization method of power transformer fault sample based on SVM SMOTE[J]. High Voltage Engineering, 2020, 46 (7): 2522- 2529.
22	杨俊闯, 赵超. K-Means聚类算法研究综述[J]. 计算机工程与应用, 2019, 55 (23): 7- 14, 63.
	YANG Junchuang, ZHAO Chao. Survey on K-means clustering algorithm[J]. Computer Engineering and Applications, 2019, 55 (23): 7- 14, 63.
23	武艺, 姚良忠, 廖思阳, 等. 一种基于改进K-means++算法的分布式光储聚合调峰方法[J]. 电网技术, 2022, 46 (10): 3923- 3931.
	WU Yi, YAO Liangzhong, LIAO Siyang, et al. A peak shaving method for distributed optical storage aggregation based on improved K-means++ algorithm[J]. Power System Technology, 2022, 46 (10): 3923- 3931.
24	HINTON G E, OSINDERO S, TEH Y W. A fast learning algorithm for deep belief nets[J]. Neural Computation, 2006, 18 (7): 1527- 1554. DOI
25	王艳, 李伟, 赵洪山, 等. 基于油中溶解气体分析的DBN-SSAELM变压器故障诊断方法[J]. 电力系统保护与控制, 2023, 51 (4): 32- 42. DOI
	WANG Yan, LI Wei, ZHAO Hongshan, et al. Fault diagnosis method of DBN-SSAELM transformer based on dissolved gas analysis in oil[J]. Power System Protection and Control, 2023, 51 (4): 32- 42. DOI
26	刘仲民, 翟玉晓, 张鑫, 等. 基于DBN-IFCM的变压器故障诊断方法[J]. 高电压技术, 2020, 46 (12): 4258- 4265.
	LIU Zhongmin, ZHAI Yuxiao, ZHANG Xin, et al. Transformer fault diagnosis method based on deep belief network and improved fuzzy C-means clustering[J]. High Voltage Engineering, 2020, 46 (12): 4258- 4265.
27	TROJOVSKÝ P, DEHGHANI M. Pelican optimization algorithm: a novel nature-inspired algorithm for engineering applications[J]. Sensors, 2022, 22 (3): 855. DOI
28	KAUR G, ARORA S. Chaotic whale optimization algorithm[J]. Journal of Computational Design and Engineering, 2018, 5 (3): 275- 284. DOI
29	吕鑫, 慕晓冬, 张钧, 等. 混沌麻雀搜索优化算法[J]. 北京航空航天大学学报, 2021, 47 (8): 1712- 1720.
	LYU Xin, MU Xiaodong, ZHANG Jun, et al. Chaos sparrow search optimization algorithm[J]. Journal of Beijing University of Aeronautics and Astronautics, 2021, 47 (8): 1712- 1720.
30	国家能源局. 变压器油中溶解气体分析和判断导则: DL/T 722—2014[S]. 北京: 中国电力出版社, 2015.
31	代杰杰, 宋辉, 杨祎, 等. 基于油中气体分析的变压器故障诊断ReLU-DBN方法[J]. 电网技术, 2018, 42 (2): 658- 664. DOI
	DAI Jiejie, SONG Hui, YANG Yi, et al. Dissolved gas analysis of insulating oil for power transformer fault diagnosis based on Re LU-DBN[J]. Power System Technology, 2018, 42 (2): 658- 664. DOI
32	李恩文. 基于重构聚类分析方法的油浸式变压器故障诊断研究[D]. 武汉: 武汉大学, 2019.
	LI Enwen. Research on fault diagnosis of oil-immersed transformer based on reconstruction cluster analysis method[D]. Wuhan: Wuhan University, 2019.
33	华丁剑. 基于支持向量机的油浸式变压器故障诊断研究[D]. 长沙: 长沙理工大学, 2012.
	HUA Dingjian. Research on fault diagnosis of oil-immersed transformer based on support vector machine[D]. Changsha: Changsha University of Science & Technology, 2012.
34	李昆鹏. 基于深度学习的油浸式变压器故障诊断方法研究[D]. 北京: 华北电力大学, 2021.
	LI Kunpeng. Research on fault diagnosis method of oil-immersed transformer based on deep learning[D]. Beijing: North China Electric Power University, 2021.
35	中国南方电网超高压输电公司组. 大型电力变压器故障诊断及案例[M]. 北京: 中国电力出版社, 2017: 164–187.
36	刘兴华. 220 kV及以下变压器故障检测典型案例分析与处理[M]. 北京: 中国电力出版社, 2018: 143–174.
37	周舟. 变压器故障色谱诊断分析[M]. 北京: 中国电力出版社, 2015: 1–4.
38	李红雷, 何清, 钱之银. 变压器油中特征气体分析诊断及检测技术[M]. 北京: 中国电力出版社, 2020: 82–83.
39	陈蕾. 电气设备故障检测诊断方法及实例[M]. 2版. 北京: 中国水利水电出版社, 2012: 40–179.
40	白雅玲, 周亚同, 刘君. 基于深度卷积嵌入聚类的日负荷曲线聚类分析[J]. 电网技术, 2022, 46 (6): 2104- 2113.
	BAI Yaling, ZHOU Yatong, LIU Jun. Clustering analysis of daily load curve based on deep convolution embedding clustering[J]. Power System Technology, 2022, 46 (6): 2104- 2113.
41	李平, 胡根铭. 基于数据增强型一维改进卷积神经网络的变压器故障诊断方法[J]. 电网技术, 2023, 47 (7): 2957- 2967.
	LI Ping, HU Genming. Transformer fault diagnosis based on data enhanced one-dimensional improved convolutional neural network[J]. Power System Technology, 2023, 47 (7): 2957- 2967.
42	荣智海, 齐波, 李成榕, 等. 面向变压器油中溶解气体分析的组合DBN诊断方法[J]. 电网技术, 2019, 43 (10): 3800- 3808.
	RONG Zhihai, QI Bo, LI Chengrong, et al. Combined DBN diagnosis method for dissolved gas analysis of power transformer oil[J]. Power System Technology, 2019, 43 (10): 3800- 3808.

函数	函数类型	维度	搜索范围	理论最优值
f₁	单峰	30	[–100, 100]ⁿ	0
f₂	多峰	30	[–5.12, 5.12]ⁿ	0

函数	函数类型	维度	搜索范围	理论最优值
f₁	单峰	30	[–100, 100]ⁿ	0
f₂	多峰	30	[–5.12, 5.12]ⁿ	0

编号	特征	编号	特征
1	H₂	8	C₂H₄/C₂H₆
2	CH₄	9	C₂H₂/(C_xH_x)
3	C₂H₆	10	H₂/(H₂+C_xH_x)
4	C₂H₄	11	C₂H₄/(C_xH_x)
5	C₂H₂	12	CH₄/(C_xH_x)
6	CH₄/H₂	13	C₂H₆/(C_xH_x)
7	C₂H₄/C₂H₂	14	(CH₄+C₂H₄)/(C_xH_x)

编号	特征	编号	特征
1	H₂	8	C₂H₄/C₂H₆
2	CH₄	9	C₂H₂/(C_xH_x)
3	C₂H₆	10	H₂/(H₂+C_xH_x)
4	C₂H₄	11	C₂H₄/(C_xH_x)
5	C₂H₂	12	CH₄/(C_xH_x)
6	CH₄/H₂	13	C₂H₆/(C_xH_x)
7	C₂H₄/C₂H₂	14	(CH₄+C₂H₄)/(C_xH_x)

类型	原样本集	训练集	测试集
N	400	320	80
PD	120	96	24
D1	110	88	22
D2	289	231	58
T1	83	66	17
T2	125	100	25
T3	379	303	76
LDT	45	36	9
HDT	44	35	9
总计	1595	1275	320

考虑样本类内不平衡的CHPOA-DBN变压器故障诊断方法

CHPOA-DBN Transformer Fault Diagnosis Method Considering Sample Within-Class Imbalance

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 14

参考文献 42

相关文章 1

编辑推荐

Metrics

过采样方法	λ_accuracy/%	λ_Kappa	λ_Macro-F1	λ_G-mean
原训练集	82.50	0.7836	0.7055	0.7405
SMOTE	85.63	0.8252	0.7366	0.8199
Borderline-SMOTE	86.56	0.8370	0.7427	0.8242
ADASYN	87.19	0.8445	0.7829	0.8521
SVM SMOTE	92.81	0.9133	0.8618	0.8866
K-means SMOTE	94.69	0.9361	0.9296	0.9420
IK-means SMOTE	96.25	0.9548	0.9502	0.9602

诊断模型	λ_accuracy/%		λ_Kappa		λ_Macro-F1		λ_G-mean
诊断模型	均衡前	均衡后	均衡前	均衡后	均衡前	均衡后	均衡前	均衡后
改良三比值	72.00	85.63	0.6661	0.8222	0.6112	0.7333	0.6692	0.8163
DBN	75.56	90.94	0.7063	0.8909	0.6486	0.8824	0.6965	0.8834
1D-CNN	76.56	92.81	0.7040	0.9135	0.6413	0.8993	0.7180	0.9013
POA-DBN	81.25	94.38	0.7707	0.9323	0.6932	0.9258	0.7252	0.9373
CHPOA-DBN	82.50	96.25	0.7836	0.9548	0.7055	0.9502	0.7405	0.9602

类型	H₂	CH₄	C₂H₆	C₂H₄	C₂H₂
N	0.6101	0.3728	0.5768	0.0488	0.0016
PD	0.9445	0.8151	0.1591	0.0244	0.0014
D1	0.6828	0.2672	0.0252	0.1682	0.5393
D2	0.5129	0.4855	0.0773	0.1739	0.2633
T1	0.3704	0.3532	0.1869	0.4577	0.0021
T2	0.1485	0.2525	0.2024	0.5088	0.0363
T3	0.0680	0.2421	0.0846	0.6724	0.0010
LDT	0.0617	0.2145	0.0423	0.0657	0.6775
HDT	0.4392	0.1123	0.0366	0.2768	0.5744

类型	H₂	CH₄	C₂H₆	C₂H₄	C₂H₂
N	0.6101	0.3728	0.5768	0.0488	0.0016
PD	0.9445	0.8151	0.1591	0.0244	0.0014
D1	0.6828	0.2672	0.0252	0.1682	0.5393
D2	0.5129	0.4855	0.0773	0.1739	0.2633
T1	0.3704	0.3532	0.1869	0.4577	0.0021
T2	0.1485	0.2525	0.2024	0.5088	0.0363
T3	0.0680	0.2421	0.0846	0.6724	0.0010
LDT	0.0617	0.2145	0.0423	0.0657	0.6775
HDT	0.4392	0.1123	0.0366	0.2768	0.5744

模态框（Modal）标题

考虑样本类内不平衡的CHPOA-DBN变压器故障诊断方法

CHPOA-DBN Transformer Fault Diagnosis Method Considering Sample Within-Class Imbalance

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 14

参考文献 42

相关文章 1

编辑推荐

Metrics