基于小波特征及深度学习的故障电弧检测  

Arc fault detection based on wavelet feature and deep learning

作　　者：余琼芳[1,2] 胡亚倩 杨艺 Yu Qiongfang;Hu Yaqian;Yang Yi(School of Electrical Engineering and Automation,Henan Polytechnic University,Jiaozuo 454003,China;Postdoctoral Programme of Beijing Research Institute,Dalian University of Technology,Beijing 100000,China)

机构地区：[1]河南理工大学电气工程与自动化学院,焦作454003 [2]大连理工大学北京研究院博士后科研工作站,北京100000

出　　处：《电子测量与仪器学报》2020年第3期100-108,共9页Journal of Electronic Measurement and Instrumentation

基　　金：国家自然科学基金(61601172);中国博士后科学基金(2018M641287)资助项目。

摘　　要：由线路绝缘层老化破损、电气接触不良等原因产生的串联故障电弧严重威胁着低压配电系统的电力安全。其电流小、温度高、隐蔽性强等特点更是给检测和识别带来了困难。基于此,提出一种基于小波特征及深度学习的串联故障电弧检测方法。通过搭建串联故障电弧实验平台,采集了典型阻性、阻感性、感性负载下的电流信号,对电流信号进行小波变换构造了训练集和测试集,通过改进的AlexNet模型识别故障电弧并输出检测结果。实验结果表明,采用该方法进行串联故障电弧识别的准确率约为95.58%,比利用AlexNet模型要高出约10.58%。Series arc fault caused by aging damage of line insulation layer and poor electrical contact seriously threatens the power safety of low voltage distribution system. It is difficult to detect and extinguish series arc fault for its characteristics of small current, high temperature and strong concealment. Because of above reasons, a method based on wavelet feature and deep learning is proposed for detecting series arc fault. Firstly, series arc fault experimental platform was built to collect the current signals under typical resistive load, inductive load and resistive-inductive load. Secondly, after transformed by wavelet transform, collected signals were decomposed to construct training sets and test sets. Finally, the arc fault was identified by the improved AlexNet model, and the test results were output. The experimental results show that the accuracy of this method for serial arc fault identification is almost 95.58%, about 10.58% higher than using AlexNet model.

关 键 词：故障电弧 小波变换 深度学习 检测方法 

分 类 号：TM501.2[电气工程—电器]