|
|
|
| Series Arc Fault Detection Method Based on Time-Domain Signal Super-Resolution |
| Yang Tianpeng1, Long Kui2, Li Qiuhui2, Mou Weiqi2, Li Bin1 |
1. State Key Laboratory of Disaster Prevention and Reduction for Power Grid Changsha University of Science and Technology Changsha 410114 China;
2. Dongguan Power Supply Bureau of Guangdong Power Grid Co. Ltd Dongguan 523129 China |
|
|
|
Abstract Series arc faults are a leading cause of electrical fires in low-voltage power distribution systems. Accurate detection of these faults is essential to ensure electrical safety, as arc faults can lead to hazardous conditions. Traditional detection methods generally rely on high-frequency fault signal features; however, these methods face challenges in practical applications due to limitations in the sensors used for fault signal acquisition. Open-type current transformers, commonly employed in real-world scenarios to ensure power supply reliability, struggle to capture high-frequency components due to their structural constraints, resulting in insufficient detection accuracy for series arc faults.++This paper proposes a novel time-domain super-resolution method for series arc fault detection, which aims to overcome the time-frequency limitations of traditional detection methods. By applying super-resolution techniques to reconstruct low-resolution data, the proposed method enhances the recovery of high-frequency components that are typically lost or obscured in traditional fault detection systems. The approach begins by setting up an arc fault simulation platform in compliance with UL1699 standards, where datasets are collected from various load types at three different sampling rates (25 kHz, 50 kHz, and 100 kHz). This enables comprehensive testing of the proposed method on data collected at varying resolutions.
A significant issue with current time-domain super-resolution algorithms is their inability to effectively restore high-frequency spectral information, which is essential for accurate arc fault detection. To address this challenge, the paper introduces a dual-domain collaborative generative time-frequency enhancement method based on TimeGAN, referred to as DCTF-GAN. This model integrates both time-domain and frequency-domain features, using a frequency-domain loss function and discriminator to improve high-frequency feature reconstruction. DCTF-GAN outperforms traditional methods and other deep learning models in terms of high-frequency detail recovery, providing a significant improvement in fault detection accuracy.
Furthermore, a lightweight RNN-Transformer hybrid detection model is developed to perform fault detection on the super-resolved data. By combining time-domain and frequency-domain features, this model achieves high detection accuracy while maintaining computational efficiency. Experimental results show that the proposed method achieves up to 99.7% detection accuracy, validating its effectiveness in various test scenarios. The results also indicate that the method significantly enhances detection performance by restoring lost high-frequency components in the signal, making it more reliable for detecting arc faults in real-world systems.
The proposed method addresses the challenges of low-frequency signal acquisition and limited high-frequency information, which have been long-standing obstacles in series arc fault detection. By leveraging the DCTF-GAN model, the system can restore high-frequency information and improve the detection capabilities of the fault detection system. Additionally, the RNN-Transformer network contributes to efficient processing, enabling real-time detection in industrial applications. This approach provides a more robust, accurate, and computationally efficient solution to series arc fault detection.
In conclusion, this paper presents a significant advancement in the detection of series arc faults. The integration of DCTF-GAN for high-frequency signal restoration and the RNN-Transformer hybrid network for fault detection represents a promising solution to the challenges of detecting arc faults in low-voltage power distribution systems. The proposed method improves detection accuracy by addressing signal acquisition limitations and enhances the overall performance of fault detection systems, making it suitable for practical deployment in electrical safety systems.
|
|
Received: 24 June 2025
|
|
|
|
|
|
[1] Thai H D, Le N B V, Lee D, et al. A survey of electrical fire causes assessment technology[J]. IEEE Access, 2024, 12: 145378-145392.
[2] 盛德杰, 王尧, 邢云琪, 等. 基于磁流体动力学仿真的低压交流串联故障电弧致火风险研究[J]. 电工技术学报, 2025, 40(10): 3326-3338.
Sheng Dejie, Wang Yao, Xing Yunqi, et al.Research on fire risk of low voltage AC series fault arc based on magnetohydrodynamics simulation[J]. Transactions of China Electrotechnical Society, 2025, 40(10): 3326-3338.
[3] 王尧, 马桐桐, 赵宇初, 等. 基于电磁辐射时延估计的串联光伏直流电弧故障定位方法[J]. 电工技术学报, 2023, 38(8): 2233-2243.
Wang Yao, Ma Tongtong, Zhao Yuchu, et al.Series DC arc-fault location method based on electromagnetic radiation delay estimation for photovoltaic systems[J]. Transactions of China Electrotechnical Society, 2023, 38(8): 2233-2243.
[4] 唐昭晖, 许志红, 叶骁勇. 交流故障电弧作用下电缆碳化的实验与仿真研究[J]. 中国电机工程学报, 2024, 44(23): 9496-9507.
Tang Zhaohui, Xu Zhihong, Ye Xiaoyong.Experi-mental and numerical investigation of cable carbonization under AC fault arc[J]. Proceedings of the CSEE, 2024, 44(23): 9496-9507.
[5] Amiri A, Samet H, Ghanbari T.Recurrence plots based method for detecting series arc faults in photovoltaic systems[J]. IEEE Transactions on Industrial Electronics, 2022, 69(6): 6308-6315.
[6] Yan Junchen, Li Qiqi, Duan Shanxu.A simplified current feature extraction and deployment method for DC series arc fault detection[J]. IEEE Transactions on Industrial Electronics, 2024, 71(1): 625-634.
[7] Yin Zhendong, Wang Li, Zhang Bin, et al.An integrated DC series arc fault detection method for different operating conditions[J]. IEEE Transactions on Industrial Electronics, 2021, 68(12): 12720-12729.
[8] Wang Yao, Hou Linming, 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.
[9] 蔡彬, 陈德桂, 吴锐, 等. 开关柜内部故障电弧的在线检测和保护装置[J]. 电工技术学报, 2005, 20(10): 83-87.
Cai Bin, Chen Degui, Wu Rui, et al.Online detecting and protection system for internal faults arc in switchgear[J]. Transactions of China Electrotechnical Society, 2005, 20(10): 83-87.
[10] 黄晨昊, 高伟. 基于超声波传感器与孤立森林的光伏系统串联电弧故障诊断方法[J]. 电气技术, 2025, 26(5): 10-16, 26.
Huang Chenhao, Gao Wei.Series arc fault diagnosis method for photovoltaic system based on ultrasonic sensor and isolation forest[J]. Electrical Engineering, 2025, 26(5): 10-16, 26.
[11] 王尧, 张彦风, 牛峰, 等. 光伏直流电弧电磁辐射特性分析与测量方法[J]. 电工技术学报, 2019, 34(14): 2913-2921.
Wang Yao, Zhang Yanfeng, Niu Feng, et al.Characterization and measurement method of DC arc electromagnetic radiation for photovoltaic systems[J]. Transactions of China Electrotechnical Society, 2019, 34(14): 2913-2921.
[12] 崔建, 孙帅, 张国钢, 等. 基于双温度磁流体电弧仿真改进Mayr电弧模型的特快速暂态过电压仿真方法[J]. 电工技术学报, 2024, 39(16): 5149-5161.
Cui Jian, Sun Shuai, Zhang Guogang, et al.The very fast transient overvoltage simulation method based on two-temperature MHD arc simulation to improve Mayr arc model[J]. Transactions of China Electro-technical Society, 2024, 39(16): 5149-5161.
[13] 刘艳丽, 郭凤仪, 李磊, 等. 一种串联型故障电弧数学模型[J]. 电工技术学报, 2019, 34(14): 2901-2912.
Liu Yanli, Guo Fengyi, Li Lei, et al.A kind of series fault arc mathematical model[J]. Transactions of China Electrotechnical Society, 2019, 34(14): 2901-2912.
[14] 王尧, 韦强强, 葛磊蛟, 等. 基于电弧电流高频分量的串联交流电弧故障检测方法[J]. 电力自动化设备, 2017, 37(7): 191-197.
Wang Yao, Wei Qiangqiang, Ge Leijiao, et al.Series AC arc fault detection based on high-frequency components of arc current[J]. Electric Power Auto-mation Equipment, 2017, 37(7): 191-197.
[15] 王尧, 李阳, 葛磊蛟, 等. 基于滑动离散傅里叶变换的串联直流电弧故障识别[J]. 电工技术学报, 2017, 32(19): 118-124.
Wang Yao, Li Yang, Ge Leijiao, et al.A series DC arcing fault recognition method based on sliding discrete Fourier transform[J]. Transactions of China Electrotechnical Society, 2017, 32(19): 118-124.
[16] 郑昕, 单潇洁. 低压交流电弧电流零区电压波形特征分析与应用[J]. 电工技术学报, 2020, 35(22): 4717-4725.
Zheng Xin, Shan Xiaojie.Characteristic analysis and application research of low voltage AC arc voltage waveform at the current zero[J]. Transactions of China Electrotechnical Society, 2020, 35(22): 4717-4725.
[17] 胡从强, 曲娜, 张帅, 等. 连续小波变换和具有注意力机制的深度残差收缩网络在低压串联电弧故障检测中的应用[J]. 电网技术, 2023, 47(5): 1897-1905.
Hu Congqiang, Qu Na, Zhang Shuai, et al.Application of continuous wavelet transform and deep residual shrinkage network with attention mechanism in detection of low voltage series arc fault[J]. Power System Technology, 2023, 47(5): 1897-1905.
[18] 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.Series arc fault detection and line selection based on local binary pattern histogram matching[J]. Transactions of China Electrotechnical Society, 2020, 35(8): 1653-1661.
[19] Jiang Run, Bao Guanghai.Series arc fault detection method based on signal-type enumeration and zoom circular convolution algorithm[J]. IEEE Transactions on Industrial Electronics, 2023, 70(10): 10607-10617.
[20] 吴建明, 杨广亮, 黄文涛. 开口式电流互感器在低压配电系统中的应用[J]. 电气技术, 2012, 13(8): 154-156.
Wu Jianming, Yang Guangliang, Huang Wentao.Application of open current transformer in low voltage distribution system[J]. Electrical Engineering, 2012, 13(8): 154-156.
[21] 林帆, 张耀, 东琦, 等. 基于分位数插值和深度自回归网络的光伏出力概率预测[J]. 电力系统自动化, 2023, 47(9): 79-87.
Lin Fan, Zhang Yao, Dong Qi, et al.Probability prediction of photovoltaic output based on quantile interpolation and deep autoregressive network[J]. Automation of Electric Power Systems, 2023, 47(9): 79-87.
[22] 程乐峰, 余涛, 张孝顺, 等. 机器学习在能源与电力系统领域的应用和展望[J]. 电力系统自动化, 2019, 43(1): 15-31.
Cheng Lefeng, Yu Tao, Zhang Xiaoshun, et al.Machine learning for energy and electric power systems: state of the art and prospects[J]. Automation of Electric Power Systems, 2019, 43(1): 15-31.
[23] Liu Guolong, Gu Jinjin, Zhao Junhua, et al.Super resolution perception for smart meter data[J]. Information Sciences, 2020, 526: 263-273.
[24] Gu J, Chen H, Liu G, et al. Super-resolution perception for industrial sensor data[J]. arXiv preprint arXiv:1809.06687, 2018.
[25] Yoon J, Jarrett D, van der Schaar M. Time-series generative adversarial networks[C]//Neural Information Processing Systems, Vancouver, 2019, 32: 5508-5518.
[26] Jiang Run, Zheng Yuesheng.Series arc fault detection using regular signals and time-series reconstruction[J]. IEEE Transactions on Industrial Electronics, 2023, 70(2): 2026-2036. |
|
|
|
2026, Vol. 41 
