Transformer Hierarchical Fault Diagnosis Model Based on Dissolved Gas Analysis of Insulating Oil and Class Overlap Features

doi:10.11930/j.issn.1004-9649.202202010

Abstract

Abstract: Dissolved gas analysis (DGA) of insulating oil can effectively identify transformer discharge fault and overheating fault. In order to improve the accuracy of transformer fault diagnosis, a transformer hierarchical fault diagnosis method is proposed based on class overlap features. Firstly, the support vector data description (SVDD) is used to divide the overlapping region of transformer fault sample data spaces, and the class overlap rate and class overlap degree are selected as the overlapping features to describe the class overlap degree and the importance of sample points respectively. And then, a hierarchical fault diagnosis model is established based on the class overlap rate. The samples of each diagnosis layer are trained separately by the separate training method, and a two-class fuzzy support vector machine (FSVM) is constructed based on class overlap degrees to diagnose faults. Experimental results show that the proposed method is more accurate than other models.

Key words: transformer fault diagnosis, class overlap, hierarchical diagnosis, SVDD, FSVM

CHEN Tie, LENG Haowei, LI Xianshan, CHEN Yifu. 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.

Add to citation manager EndNote|Ris|BibTeX

URL: https://www.electricpower.com.cn/EN/10.11930/j.issn.1004-9649.202202010

https://www.electricpower.com.cn/EN/Y2022/V55/I7/22

References

[1] 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
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
[2] 李恩文, 王力农, 宋斌, 等. 基于混沌序列的变压器油色谱数据并行聚类分析[J]. 电工技术学报, 2019, 34(24): 5104–5114
LI Enwen, WANG Linong, SONG Bin, et al. Parallel clustering analysis of dissolved gas analysis data based on chaotic sequences full text replacement[J]. Transactions of China Electrotechnical Society, 2019, 34(24): 5104–5114
[3] 王文博. 基于不平衡数据集的变压器缺陷预测[D]. 北京: 华北电力大学, 2020.
WANG Wenbo. Transformer defect prediction based on imbalanced data set[D]. Beijing: North China Electric Power University, 2020.
[4] 崔宇, 侯慧娟, 苏磊, 等. 考虑不平衡案例样本的电力变压器故障诊断方法[J]. 高电压技术, 2020, 46(1): 33–41
CUI Yu, HOU Huijuan, SU Lei, et al. Fault diagnosis method for power transformer considering imbalanced class distribution[J]. High Voltage Engineering, 2020, 46(1): 33–41
[5] 田臣, 周丽娟. 基于带多数类权重的少数类过采样技术和随机森林的信用评估方法[J]. 计算机应用, 2019, 39(6): 1707–1712
TIAN Chen, ZHOU Lijuan. Credit assessment method based on majority weight minority oversampling technique and random forest[J]. Journal of Computer Applications, 2019, 39(6): 1707–1712
[6] KIM S, PARK J, KIM W, et al. Learning from even a weak teacher: Bridging rule-based Duval method and a deep neural network for power transformer fault diagnosis[J]. International Journal of Electrical Power & Energy Systems, 2022, 136: 107619.
[7] 张天翼, 丁立新. 一种基于SMOTE的不平衡数据集重采样方法[J]. 计算机应用与软件, 2021, 38(9): 273–279
ZHANG Tianyi, DING Lixin. A new resampling method based on smote for imbalanced data set[J]. Computer Applications and Software, 2021, 38(9): 273–279
[8] 王忠震, 黄勃, 方志军, 等. 改进SMOTE的不平衡数据集成分类算法[J]. 计算机应用, 2019, 39(9): 2591–2596
WANG Zhongzhen, HUANG Bo, FANG Zhijun, et al. Improved SMOTE unbalanced data integration classification algorithm[J]. Journal of Computer Applications, 2019, 39(9): 2591–2596
[9] 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
[10] 董宏成, 文志云, 万玉辉, 等. 基于DPC聚类重采样结合ELM的不平衡数据分类算法[J]. 计算机工程与科学, 2021, 43(10): 1856–1863
DONG Hongcheng, WEN Zhiyun, WAN Yuhui, et al. An imbalanced data classification algorithm based on DPC clustering resampling combined with ELM[J]. Computer Engineering & Science, 2021, 43(10): 1856–1863
[11] 李黄曼, 张勇, 张瑶. 基于ISSA优化SVM的变压器故障诊断研究[J]. 电子测量与仪器学报, 2021, 35(3): 123–129
LI Huangman, ZHANG Yong, ZHANG Yao. Study of transformer fault diagnosis based on improved sparrow search algorithm optimized support vector machine[J]. Journal of Electronic Measurement and Instrumentation, 2021, 35(3): 123–129
[12] 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
[13] 黄新波, 王享, 田毅, 等. 基于PSO-ELM融合动态加权AdaBoost的变压器故障诊断方法[J]. 高压电器, 2020, 56(5): 39–46
HUANG Xinbo, WANG Xiang, TIAN Yi, et al. Transformer fault diagnosis algorithm based on PSO-ELM fusion dynamically weighted AdaBoost[J]. High Voltage Apparatus, 2020, 56(5): 39–46
[14] 熊海涛, 吴俊杰, 刘洪甫, 等. 分类中的类重叠问题及其处理方法研究[J]. 管理科学学报, 2013, 16(4): 8-21.
XIONG Haitao, WU Junjie, LIU Hongfu, et al Research on class overlap in classification and its processing methods [J] Journal of management science in China, 2013, 16(4): 8-21.
[15] VUTTIPITTAYAMONGKOL P, ELYAN E, PETROVSKI A. On the class overlap problem in imbalanced data classification[J]. Knowledge-Based Systems, 2021, 212(1): 106631.
[16] LIU C L. Partial discriminative training for classification of overlapping classes in document analysis[J]. International Journal of Document Analysis and Recognition (IJDAR), 2008, 11(2): 53–65.
[17] GARCÍA V, ALEJO R, SÁNCHEZ J S, et al. Combined effects of class imbalance and class overlap on instance-based classification[C]//Intelligent Data Engineering and Automated Learning – IDEAL 2006, 2006.
[18] STEFANOWSKI J. Overlapping, rare examples and class decomposition in learning classifiers from imbalanced data[M]. Springer Berlin Heidelberg, 2013.
[19] VUTTIPITTAYAMONGKOL P, ELYAN E. Neighbourhood-based under sampling approach for handling imbalanced and overlapped data[J]. Information Sciences, 2020, 509(1): 47–70.
[20] 吴园园, 申立勇. 基于类重叠度欠采样的不平衡模糊多类支持向量机[J]. 中国科学院大学学报, 2018, 35(4): 536–543
WU Yuanyuan, SHEN Liyong. Imbalanced fuzzy multiclass support vector machine algorithm based on class-overlap degree undersampling[J]. Journal of University of Chinese Academy of Sciences, 2018, 35(4): 536–543
[21] 刘晟, 朱玉全, 孙金津. 基于核空间相对密度的SVDD多类分类算法[J]. 计算机应用研究, 2010, 27(5): 1694–1696
LIU Sheng, ZHU Yuquan, SUN Jinjin. SVDD multiclass classification algorithm based on relative density in kernel space[J]. Application Research of Computers, 2010, 27(5): 1694–1696
[22] GARCíA V, MOLLINEDA R A, SáNCHEZ J S. On the k-NN performance in a challenging scenario of imbalance and overlapping[J]. Pattern Analysis and Applications, 2008, 11(3/4): 269–280.
[23] 瞿俊. 基于重叠度的层次聚类算法研究及其应用[D]. 厦门: 厦门大学, 2007.
QU Jun. Research on overlap similarity-based hierarchical clustering algorithms and its application[D]. Xiamen: Xiamen University, 2007.
[24] 李鹏, 黄培炜, 丁瀛, 等. 基于NSGA-Ⅱ和SVDD的转向架构架异常状态监测[J]. 传感技术学报, 2021, 34(7): 874–879
LI Peng, HUANG Peiwei, DING Ying, et al. Abnormal condition monitoring of bogie frame based on NSGA-Ⅱ and SVDD[J]. Chinese Journal of Sensors and Actuators, 2021, 34(7): 874–879
[25] 李凯, 李洁. 基于pinball损失的结构模糊多分类支持向量机算法[J]. 计算机应用, 2021, 41(11): 3104–3112
LI Kai, LI Jie. Structure-fuzzy multi-class support vector machine algorithm based on pinball loss[J]. Journal of Computer Applications, 2021, 41(11): 3104–3112
[26] 韩赛赛, 刘宝柱, 艾欣. 基于MCMC方法和油色谱数据的变压器动态故障率模型[J]. 电力系统保护与控制, 2019, 47(15): 1–8
HAN Saisai, LIU Baozhu, AI Xin. Transformer dynamic failure rate model based on MCMC method and oil chromatographic data[J]. Power System Protection and Control, 2019, 47(15): 1–8
[27] 方涛, 钱晔, 郭灿杰, 等. 基于天牛须搜索优化支持向量机的变压器故障诊断研究[J]. 电力系统保护与控制, 2020, 48(20): 90–96
FANG Tao, QIAN Ye, GUO Canjie, et al. Research on transformer fault diagnosis based on a beetle antennae search optimized support vector machine[J]. Power System Protection and Control, 2020, 48(20): 90–96
[28] 张弛, 吴东, 王伟, 等. 不平衡样本下基于变分自编码器预处理深度学习和DGA的变压器故障诊断方法[J]. 南方电网技术, 2021, 15(3): 68–74
ZHANG Chi, WU Dong, WANG Wei, et al. Transformer fault diagnosis method based on variational auto-encoders preprocessing deep learning and DGA for unbalanced samples[J]. Southern Power System Technology, 2021, 15(3): 68–74