采用高频特性的低压电弧故障识别方法  

Method of Low-voltage Arc Fault Recognition Using High Frequency Feature

作　　者：高艳艳[1] 张认成[1] 杨建红[1] 杜建华[1] 杨凯[1] 

机构地区：[1]华侨大学机电及自动化学院,厦门361021

出　　处：《电力系统及其自动化学报》2016年第6期49-55,共7页Proceedings of the CSU-EPSA

基　　金：福建省高校产学合作科技计划重大项目(2012H6013);福建省科技计划重点项目(2013H0028);福建省自然科学基金资助项目(2012J01214)

摘　　要：为准确识别交流电力系统中的电弧故障问题,针对不同类型负载的电弧故障,提出一种基于小波熵的电弧故障普适性检测方法。运用小波变换提取电弧故障发生时在电流过零点附近产生的高频信号,采用该高频信号的小波熵表征电弧故障的突变信息,并利用最小二乘支持向量机对小波熵进行分类,实现对电弧故障的有效识别。结果表明,在文中的实验条件下能够全部识别出电弧故障。该方法不仅可以对单一负载和组合型负载的电弧故障进行识别,还可以避免负载正常电弧和负载启动过程引起的误判,也能克服一些抑制性负载的干扰。In order to accurately identify the arc fault in AC system,for different types of load,a universal arc fault de-tection method based on wavelet entropy was proposed. This paper discussed the high frequency signals features whenthe arc occurs,the wavelet transforming are applied to extract the high frequency spectrum of the arc fault signals andused its wavelet energy entropy as feature parameters,then the arc fault was effectively identified by using the leastsquares support vector machine（SVM）where wavelet entropy were classified. Identification results show that arc faultwas entirely recognized under the experiment conditions,this method can not only identify arc fault of the single andcombination loads,but also avoid some miscalculations which some loads brought out under the processing of startup orworking and can also overcome disturbances which generated by some inhibitory loads.

关 键 词：电弧故障 小波熵 高频辐射 支持向量机 小波变换 

分 类 号：TM51[电气工程—电器]