基于FastICA的低压电弧特征分析  

Feature Analysis of Arc Fault in Low Voltage Based on FastICA

作　　者：李夏河[1] 张认成[1] 杨建红[1] 

机构地区：[1]华侨大学机电及自动化学院,福建厦门361021

出　　处：《低压电器》2012年第10期4-7,46,共5页Low Voltage Apparatus

基　　金：福建省科技计划重大项目(2012H6013);福建省自然科学基金(2012J01214)

摘　　要：针对电弧故障特征量提取难的问题,将独立分量分析方法引入到电弧故障特征提取研究领域。依照UL 1699标准搭建电弧故障试验平台,采集试验电路中的电压、电流信号,采用FastICA算法对试验数据进行分离。通过不同负载分离结果的对比分析,可提取电弧故障的典型特征。试验结果表明,该方法能够成功获取故障信号的各独立分量,进而更有效地提取电弧故障的典型特征。As it is difficult to extract arc fault feature vectors, an independent component analysis method was introduced to arc fauh features extraction. An arc fault experimental platform was built based on UL 1699, the current and voltage signals were colleted in the experiences. Using FastICA algorithm of independent component analysis to separate the waveform, typical characteristics of arc fault would be extracted according to the comparative analysis of separated results . The results show that this method can obtain each component from arc fault data successfully, thus typical features of arc fault can be extracted more effectively.

关 键 词：电弧故障 独立分量分析 傅里叶级数 谐波含有率 

分 类 号：TM501.2[电气工程—电器]