电气工程学报 ›› 2016, Vol. 11 ›› Issue (8): 23-29.doi: 10.11985/2016.08.004

• • 上一篇    下一篇

基于K均值聚类的串联型故障电弧识别

陈昌垦,刘艳丽,李颖   

  1. 辽宁工程技术大学电气与控制工程学院 葫芦岛 125105
  • 收稿日期:2016-02-21 出版日期:2016-08-25 发布日期:2016-08-25
  • 作者简介:陈昌垦,男 1992年生,硕士研究生,研究方向为电力系统安全与故障检测。|刘艳丽,女 1981年生,博士研究生,讲师,研究方向为煤矿电气安全。
  • 基金资助:
    国家自然科学基金(51277090);辽宁省教育厅重点实验室基础研究项目(LZ2014024);辽宁工程技术大学第三批生产技术问题创新研究基金资助(14-T-004)

Recognition of Series Arc Fault Based on K-Means Clustering

Changken Chen,Yanli Liu,Ying Li   

  1. Liaoning Technical University Huludao 125105 China
  • Received:2016-02-21 Online:2016-08-25 Published:2016-08-25

摘要:

针对供配电系统中经常出现串联型故障电弧引发火灾等事故的问题,研制了串联型故障电弧实验装置,并针对典型负载开展了大量实验。首先,利用傅里叶变换提取了发生故障电弧前后电流的前20次谐波含量,并将其作为样本;其次,采用主成分分析对样本数据进行降维,提取出电流谐波变化的主要成分;最后,运用K均值聚类判断出原始信号是否故障。结果表明,以电流谐波为特征,通过主成分分析和K均值聚类可以有效地识别串联型故障电弧。

关键词: 串联型故障电弧, 谐波含量, 主成分分析, K均值聚类

Abstract:

In view of the fire and other accidents caused by series arc fault in a power supply and distribution system, an experimental device of series arc fault is developed and a lots of experiments about typical loads are carried out. Firstly, the first 20 times harmonic content of the current before and after arc fault are extracted by Fourier transform. Secondly, the dimension of sample data is reduced by using principal component analysis(PCA). A few main components are extracted to reflect the fluctuations of current signal. Finally, the main harmonic data are analyzed by K-means clustering to determine whether the fault occurred in original signal. The results show that the series arc fault could be recognized through principal component analysis and K-means clustering.

Key words: Series arc fault, harmonic content, principal component analysis, K-means clustering

中图分类号: