基于分布式光纤传感系统的有载变压器故障检测技术研究

doi:10.3976/j.issn.1002-4026.2023.05.007

Abstract

Abstract:

This paper proposes an artificial neural network-based fault detection and prediction model for on-load transformers using distributed fiber optic sensing technology. By artificially simulating the fault and normal operating states of transformers and using the k-means synthetic minority oversampling technique data expansion method, a small number of fault datasets can be limitedly expanded so that the amount of fault data can be consistent with that of normal data. Therefore, the expanded fault data and normal operation data can be input into the convolutional neural networks long short term memory identification model. Finally, the fault recognition rate can be increased to 100%, which has significant implications for the development of fault recognition systems for on-load transformers based on distributed fiber optic sensing technology.

Key words: neural network, distributed acoustic sensor, on-load transformer, fault detection, smote, SMOTE algorithm, Pattern recognition, data enhancement

CLC Number:

TN247

DONG Guanlei, JIANG Xiaodong, SUN Peng, YANG Guang, GENG Junqi, WANG Jiawen, QU Shuai, HUANG Sheng, WANG Chen, SHANG Ying. On-load transformer fault detection based on distributed optical fiber sensing system[J].Shandong Science, 2023, 36(5): 52-59.

Figures/Tables 7

Fig.1

Fig.2

Fig.3

Table 1

Fig.4

Fig.5

Table 2

References 19

[1]	林明霞. 一种实现φ-OTDR系统长距离监测的方法及系统:CN201410817348.X[P]. 2015-03-25.
[2]	张知先, 陈伟根, 汤思蕊, 等. 基于互补集总经验模态分解和局部异常因子的有载分接开关状态特征提取及异常状态诊断[J]. 电工技术学报, 2019, 34(21): 4508-4518. DOI: 10.19595/j.cnki.1000-6753.tces.181320.
[3]	聂鹏. 浅谈电力变压器故障分析及其诊断方法[J]. 科技创新与应用, 2014(12):121.
[4]	石秀立. 基于振动信号特征的有载分接开关机械故障诊断策略研究[D]. 焦作: 河南理工大学, 2018.
[5]	钟龙申, 高学军, 王振友. 一种新的基于K-means改进的SMOTE算法在不平衡数据分类中的应用[J]. 数学的实践与认识, 2015, 45(19):198-206.
[6]	赵彤. 有载分接开关机械状态的在线监测与故障诊断技术研究[D]. 济南: 山东大学, 2008.
[7]	SHI Y, WANG Y Y, WANG L Y, et al. Multi-event classification for Φ-OTDR distributed optical fiber sensing system using deep learning and support vector machine[J]. Optik, 2020, 221: 165373. DOI: 10.1016/j.ijleo.2020.165373.
[8]	SUN Q, FENG H, YAN X Y, et al. Recognition of a phase-sensitivity OTDR sensing system based on morphologic feature extraction[J]. Sensors (Basel, Switzerland), 2015, 15(7): 15179-15197. DOI: 10.3390/s150715179.
[9]	孙钢. 基于SVM的入侵检测系统研究[D]. 北京: 北京邮电大学, 2007.
[10]	JIA H Z, LIANG S, LOU S Q, et al. A k-nearest neighbor algorithm-based near category support vector machine method for event identification of φ-OTDR[J]. IEEE Sensors Journal, 2019, 19(10): 3683-3689. DOI: 10.1109/JSEN.2019.2891750.
[11]	ZHANG T, LI Z N, ZHU Q, et al. Improved procedures for training primal Wasserstein GANs[C]//2019 IEEE SmartWorld, Ubiquitous Intelligence & Computing, Advanced & Trusted Computing, Scalable Computing & Communications, Cloud & Big Data Computing, Internet of People and Smart City Innovation (SmartWorld/SCALCOM/UIC/ATC/CBDCom/IOP/SCI). Leicester, UK. IEEE, 2020: 1601-1607. DOI: 10.1109/SmartWorld-UIC-ATC-SCALCOM-IOP-SCI.2019.00286.
[12]	WU H J, CHEN J P, LIU X R, et al. One-Dimensional CNN-Based Intelligent Recognition of Vibrations in Pipeline Monitoring With DAS[J]. Journal of Lightwave Technology, 2019, 37(17): 4359-4366. DOI:10.1109/JLT.2019.2923839
[13]	SNELL J, SWERSKY K, ZEMEL R S. Prototypical networks for few-shot learning[EB/OL]. 2017: arXiv: 1703.05175. https://arxiv.org/abs/1703.05175.
[14]	吴飞. 人工智能: 从规则学习, 到数据驱动, 再到能力增强[J]. 杭州科技, 2017(2): 51-53.
[15]	MA G M, SHI C, QIN W Q, et al. A non-intrusive electrical discharge localization method for gas insulated line based on phase-sensitive OTDR and Michelson interferometer[J]. IEEE Transactions on Power Delivery, 2019, 34(4): 1324-1331. DOI: 10.1109/TPWRD.2019.2892306.
[16]	WU L W, HUANG S S, WU F, et al. Protein subnuclear localization based on radius-SMOTE and kernel linear discriminant analysis combined with random forest[J]. Electronics, 2020, 9(10): 1566. DOI: 10.3390/electronics9101566.
[17]	张俊楠, 娄淑琴, 梁生. 基于SVM算法的φ-OTDR分布式光纤扰动传感系统模式识别研究[J]. 红外与激光工程, 2017, 46(4): 219-225. DOI: 10.3788/IRLA201746.0422003.
[18]	刘明山, 于艳玲, 周原, 等. 基于k-means改进的SMOTE算法: CN110852388A[P]. 2020-02-28.
[19]	沈艳, 余冬华, 王昊雷. 粒子群k-means聚类算法的改进[J]. 计算机工程与应用, 2014, 50(21): 125-128. DOI: 10.3778/j.issn.1002-8331.1401-0357.

类别	预测错误(false)	预测准确(true)
反例	T_N	F_N
正例	F_P	T_P

故障指标	无k-means SMOTE的ANN	经过k-means SMOTE的ANN	无k-means SMOTE的CNN-LSTM	经过k-means SMOTE的CNN-LSTM
准确率	0.998 2	0.996 0	0.997 0	1.000 0
精确率	0.500 0	1.000 0	0.916 6	1.000 0
召回率	0.750 0	0.888 8	1.000 0	1.000 0
F₁值	0.600	0.941	0.950	1.000

On-load transformer fault detection based on distributed optical fiber sensing system

RichHTML

PDF (PC)

Like

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 7

References 19

Related Articles 15

Metrics

Comments

Recommended 0

[1]	XING Hao, ZHANG Ruiyun, XU Tengfei, JI Feng. Prediction and analysis of fabric drape coefficient based on genetic-algorithm optimized BP neural network [J]. Shandong Science, 2025, 38(1): 53-63.
[2]	ZHANG Hanxiao, LIU Yuran, LIU Yuan, NIU Zichen. A multitask learning model for the prediction of short-term subway passenger flow [J]. Shandong Science, 2024, 37(1): 95-106.
[3]	JIANG Xiaodong, WANG Leilei, SUN Peng, YANG Guang, GENG Junqi, WANG Jiawen, HUANG Sheng, QU Shuai, WANG Chen, SHANG Ying. Fault detection of an on-load tap changer based on generative adversarial network [J]. Shandong Science, 2023, 36(6): 68-73.
[4]	ZHUANG Hao, LI Yang, TAO Mingkun. CNN-LSTM driver intention recognition method based on Attention mechanism [J]. Shandong Science, 2023, 36(2): 103-111.
[5]	ZENG Zhao-hui,WANG Jiang-feng,JIAO Xin-ping,XIONG Hui-yuan,GONG Xi-zhi. Highway traffic state recognition based on swarm intelligence [J]. Shandong Science, 2022, 35(3): 72-81.
[6]	SUN Feng-juan,TIAN Yong,SUN Kai-zheng,FU Hua-xuan,ZHANG Wen-juan,LI Min,L&xDC; Chen. Ozone prediction in Jinan based on artificial neural network ensemble prediction [J]. Shandong Science, 2022, 35(3): 89-99.
[7]	LI Lian-wei,ZHANG Yuan-yu,YUE Zeng-you,XUE Cun-jin,FU Yu-xuan,XU Yang-feng. Extracting inland cage aquacultural areas from high-resolution remote sensing images using fully convolutional networks model [J]. Shandong Science, 2022, 35(2): 1-10.
[8]	CAO Lu-hui, DENG Yu-xiang, CHEN Tong, LI Zhao. A deep learning-based method for Chinese text-feature extraction and classification [J]. Shandong Science, 2019, 32(6): 106-111.
[9]	WANG Shuo, GU Yuan-li, LI Meng, LU Wen-qi, ZHANG Yuan. Short term traffic flow prediction of expressway based on chaos theory and MEA-BPNN model [J]. Shandong Science, 2019, 32(2): 98-107.
[10]	LI Xiao-feng，YU Hui-bin. Research on intelligent PID control of ship course based on BP neural network [J]. SHANDONG SCIENCE, 2018, 31(4): 8-14.
[11]	ZHENG Wei,YANG Ying,HUI Li, LU Cheng-jie,ZHAO Bin,YANG Li. The analysis between wave estimation error and influencing factors based on BP model when ADCP tilts [J]. SHANDONG SCIENCE, 2018, 31(3): 1-7.
[12]	HUANG En-tan, GU Yuan-li. An optimized wavelet neural network based on improved artificial bee colony algorithm for short-term traffic flow prediction [J]. SHANDONG SCIENCE, 2018, 31(2): 79-87.
[13]	LIU Dongyan, ZHANG Yingying, WU Bingwei, ZHANG Ying, HU yunchuan. Simulation analysis and control of solid oxide fuel cell system [J]. SHANDONG SCIENCE, 2016, 29(1): 1-6.
[14]	QIN Jian,HE Shiwei. BP neural network based volume forecast of wholly social containerized freight [J]. SHANDONG SCIENCE, 2015, 28(2): 70-74.
[15]	LIU Fei, GU Wenjing, WANG Gongtang, WAN Honglin. BP neural network based classification method for iris image quality [J]. SHANDONG SCIENCE, 2015, 28(2): 108-112.