多负载回路串联型故障电弧智能诊断及选线方法研究*

doi:10.11985/2023.04.040

摘要/Abstract

摘要：

低压用电系统线路结构复杂，负载形式多样，因线路老化、接触不良等原因发生串联型故障电弧时，现有的故障电弧断路器通常只能对单相单负载回路进行保护，难以检测到下级线路发生的故障。因此，在AC 220 V下搭建了家用多负载回路串联型故障电弧试验平台，开展各支路及干路故障下的故障电弧试验，获得不同工况下发生故障电弧时的干路电流数据，并将其作为样本建立模型样本库来训练建立的一维卷积神经网络。经过大量训练，建立的家用多负载回路串联型故障电弧诊断及选线模型在验证集上的准确率达到了98.6%。随后使用测试集中的连续电流数据对模型的鲁棒性、抗扰性及泛化性进行测试，最后将训练好的模型导入树莓派中进行运行时间测试，结果表明其检测时间满足UL标准的要求，但在投入实际应用前，仍需要对模型做进一步优化使得其更适合在嵌入式系统上运行。

关键词: 串联型故障电弧, 深度学习, 一维卷积神经网络, 故障诊断, 故障选线

Abstract:

The circuit structure of the low-voltage power system is complex, and the load forms are various. When a series arc fault occurs due to line aging, poor contact, etc., the existing arc fault circuit breaker can usually only protect the single-phase single-load circuit, and it is difficult to detect the lower level line failure. Therefore, a household multi-load circuit series arc fault experiment platform is set up under AC 220 V to carry out arc fault experiments under various branch and trunk faults, and obtain the trunk current data when fault arcs occurred under different working conditions, which are used as samples to establish a model sample database to train the established one-dimensional convolutional neural network. After a lot of training, the established household multi-load circuit series arc fault diagnosis and line selection model has an accuracy rate of 98.6% on the validation set. Then, the continuous current data in the test set is used to verify the robustness, immunity and generalization of the model, and finally the trained model is imported into the Raspberry Pi for running time test. The results show that the detection time meets the requirements of the UL standard. However, before it is put into practice, the model still needs to be further optimized to make it more suitable to run on embedded system.

Key words: Series fault arc, deep learning, one-dimensional convolutional neural network, fault diagnosis, fault line selection

中图分类号:

TM561

刘艳丽, 张晓乐, 吕正阳, 徐振豪, 刘洋. 多负载回路串联型故障电弧智能诊断及选线方法研究^*[J]. 电气工程学报, 2023, 18(4): 378-388.

LIU Yanli, ZHANG Xiaole, LÜ Zhengyang, XU Zhenhao, LIU Yang. Research on Intelligent Diagnosis and Route Selection Method of Multi-load Circuit Series Fault Arc[J]. Journal of Electrical Engineering, 2023, 18(4): 378-388.

图/表 20

图1

图2

表1

表2

图3

图4

图5

图6

图7

表3

表4

图8

图9

图10

图11

图12

图13

图14

图15

表5

参考文献 21

[1]	崔芮华, 王洋, 王传宇, 等. 基于多信息融合的航空线路串联故障电弧识别方法[J]. 电工技术学报, 2019, 34(S1):118-125.
	CUI Ruihua, WANG Yang, WANG Chuanyu, et al. A method for recognizing series arc fault of aviation lines based on multi-information fusion[J]. Transactions of China Electrotechnical Society, 2019, 34(S1):118-125.
[2]	郭凤仪, 高洪鑫, 唐爱霞, 等. 局部二值模式直方图匹配的串联故障电弧检测及选线[J]. 电工技术学报, 2020, 35(8):1653-1661.
	GUO Fengyi, GAO Hongxin, TANG Aixia, et al. Detectionand line selection of series fault arc based on local binary mode histogram matching[J]. Transactions of China Electrotechnical Society, 2020, 35(8):1653-1661.
[3]	刘丽智, 刘艳丽, 殷煜炜, 等. 基于关联维数与小波包-AR谱的串联故障电弧特征研究[J]. 电子测量与仪器学报, 2018, 32(9):108-116.
	LIU Lizhi, LIU Yanli, YIN Yuwei, et al. Research on series fault arc characteristics based on correlation dimension and wavelet packet-AR spectrum[J]. Journal of Electronic Measurement and Instrument, 2018, 32(9):108-116.
[4]	王毅, 陈进, 李松浓, 等. 基于时频域分析和随机森林的故障电弧检测[J]. 电子测量与仪器学报, 2021, 35(5):62-68.
	WANG Yi, CHEN Jin, LI Songnong, et al. Fault arc detection based on time-frequency domain analysis and random forest[J]. Journal of Electronic Measurement and Instrument, 2021, 35(5):62-68.
[5]	谢国民, 刘宽. 基于形态学滤波和OTSU的串联故障电弧识别方法[J]. 电子测量与仪器学报, 2019, 33(5):46-56.
	XIE Guomin, LIU Kuan. A method of series fault arc recognition based on morphological filtering and OSTU[J]. Journal of Electronic Measurement and Instrument, 2019, 33(5):46-56.
[6]	曾珂, 邢丽冬, 高杨, 等. 变频系统中交流电弧故障特性及检测技术[J]. 南京航空航天大学学报, 2020, 52(2):215-223.
	ZENG Ke, XING Lidong, GAO Yang, et al. AC arc fault characteristics and detection technology in frequency conversion system[J]. Journal of Nanjing University of Aeronautics and Astronautics, 2020, 52(2):215-223.
[7]	赵怀军, 秦海燕, 刘凯, 等. 基于相关理论及零休特征融合的串联故障电弧检测方法[J]. 仪器仪表学报, 2020, 41(4):218-228.
	ZHAO Huaijun, QIN Haiyan, LIU Kai, et al. A series fault arc detection method based on related theories and zero-rest feature fusion[J]. Chinese Journal of Scientific Instrument, 2020, 41(4):218-228.
[8]	唐圣学, 刁旭东, 陈丽, 等. 光伏发电系统直流串联微弱故障电弧检测方法研究[J]. 仪器仪表学报, 2021, 42(3):150-160.
	TANG Shengxue, DIAO Xudong, CHEN Li, et al. Research on detection method of weak fault arc in DC series of photovoltaic power generation system[J]. Chinese Journal of Scientific Instrument, 2021, 42(3):150-160.
[9]	郭凤仪, 邓勇, 王智勇, 等. 基于灰度-梯度共生矩阵的串联故障电弧特征[J]. 电工技术学报, 2018, 33(1):71-81.
	GUO Fengyi, DENG Yong, WANG Zhiyong, et al. Characteristics of series fault arc based on gray-gradient co-occurrence matrix[J]. Transactions of China Electrotechnical Society, 2018, 33(1):71-81.
[10]	余琼芳, 胡亚倩, 杨艺. 基于小波特征及深度学习的故障电弧检测[J]. 电子测量与仪器学报, 2020, 34(3):100-108.
	YU Qiongfang, HU Yaqian, YANG Yi. Arc fault detection based on wavelet features and deep learning[J]. Journal of Electronic Measurement and Instrument, 2020, 34(3):100-108.
[11]	褚若波, 张认成, 杨凯, 等. 基于多层卷积神经网络的串联电弧故障检测方法[J]. 电网技术, 2020, 44(12):4792-4798.
	CHU Ruobo, ZHANG Rencheng, YANG Kai, et al. Series arc fault detection method based on multi-layer convolutional neural network[J]. Power System Technology, 2020, 44(12):4792-4798.
[12]	余琼芳, 黄高路, 杨艺, 等. 基于AlexNet深度学习网络的串联故障电弧检测方法[J]. 电子测量与仪器学报, 2019, 33(3):145-152.
	YU Qiongfang, HUANG Gaolu, YANG Yi, et al. Series fault arc detection method based on AlexNet deep learning network[J]. Journal of Electronic Measurement and Instrument, 2019, 33(3):145-152.
[13]	WANG Y, HOU L, PAUL K C, et al. Arcnet:Series AC arc fault detection based on raw current and convolutional neural network[J]. IEEE Transactions on Industrial Informatics, 2022, 18(1):77-86. doi: 10.1109/TII.2021.3069849
[14]	CHU R, ZHANG R, HUANG Q, et al. TDV-LSTM:A new methodology for series arc fault detection in low power ac systems[C]// 2020 IEEE Sustainable Power and Energy Conference (iSPEC), 2020:2319-2324.
[15]	HAN X, LI D, HUANG L, et al. Series arc fault detection method based on category recognition and artificial neural network[J]. Electronics, 2020, 9(9):1367. doi: 10.3390/electronics9091367
[16]	熊庆, 陈维江, 汲胜昌, 等. 低压直流系统故障电弧特性、检测和定位方法研究进展综述[J]. 中国电机工程学报, 2020, 40(18):6015-6027.
	XIONG Qing, CHEN Weijiang, JI Shengchang, et al. Summary of research progress on characteristics,detection and location methods of arc fault in low voltage DC system[J]. Proceedings of the CSEE, 2020, 40(18):6015-6027.
[17]	白辉, 许志红. 基于小波包变换和高阶累积量的电弧故障识别方法[J]. 电力自动化设备, 2020, 40(11):195-202.
	BAI Hui, XU Zhihong. Arc fault identification method based on wavelet packet transform and high-order cumulant[J]. Electric Power Automation Equipment, 2020, 40(11):195-202.
[18]	PAN J, ZI Y, CHEN J, et al. LiftingNet:A novel deep learning network with layerwise feature learning from noisy mechanical data for fault classification[J]. IEEE Transactions on Industrial Electronics, 2018, 65(6):4973-4982. doi: 10.1109/TIE.41
[19]	SOHAIB M, KIM J M. A robust deep learning based fault diagnosis of rotary machine bearings[J]. Journal of Computational and Theoretical Nanoscience, 2017, 23(12):12797-12801.
[20]	喻后聃, 米秋实, 赵栋, 等. 基于一维卷积神经网络的光纤周界入侵模式识别[J]. 光子学报, 2021, 50(9):95-105.
	YU Houdan, MI Qiushi, ZHAO Dong, et al. Pattern recognition of fiber perimeter intrusion based on one-dimensional convolutional neural network[J]. Acta Photonica Sinica, 2021, 50(9):95-105.
[21]	李美, 王一玮, 李林, 等. 绝缘气体对开关柜内部故障燃弧压力上升的影响[J]. 电气工程学报, 2020, 15(2):11-17.
	LI Mei, WANG Yiwei, LI Lin, et al. The influence of insulating gas on the rise of arcing pressure in the switchgear’s internal fault[J]. Journal of Electrical Engineering, 2020, 15(2):11-17.

负载名称	运行功率/W	负载名称	运行功率/W
灯泡	200	手电钻	830
小太阳	600	洗衣机	200
热水壶	500	吹风机	2 000
电磁炉	1 000

试验组别	回路数	故障点	负载
1~6	5	1、2、4、5、6、8	灯泡、小太阳、电磁炉、手电钻、洗衣机
7~12	5	1、2、3、4、7、8	灯泡、小太阳、热水壶、电磁炉、吹风机

故障类型	标签值
灯泡故障	1
电磁炉故障	2
干路故障	3
手电钻故障	4
洗衣机故障	5
小太阳故障	6
正常状态	0

层类型	输出神经元	参数量
卷积(32,11)	(None,2000,32)	384
卷积(32,11)	(None,2000,32)	11 296
最大池化	(None,666,32)	0
Dropout	(None,666,32)	0
卷积(64,11)	(None,666,64)	22 592
卷积(64,11)	(None,666,64)	45 120
最大池化	(None,222,64)	0
Dropout	(None,222,64)	0
卷积(128,11)	(None,222,128)	90 240
卷积(128,11)	(None,222,128)	180 352
最大池化	(None,74,128)	0
Dropuut	(None,74,128)	0
卷积(256,11)	(None,74,256)	360 704
卷积(256,11)	(None,74,256)	721 152
Dropout	(None,74,256)	0
平均池化	(None,256)	0
Dropout	(None,256)	0
全连接	(None,7)	1 799

方法及文献	检测准确率(%)	能否实现选线	采样率/ kHz
TDV-CNN ^[11]	97	否	1 000
AlexNet ^[12]	85	否	50
ArcNet ^[13]	99.05	否	10
本文所提1D-CNN	98.60	能	100