电器与能效管理技术 ›› 2023, Vol. 0 ›› Issue (4): 1-7.doi: 10.16628/j.cnki.2095-8188.2023.04.001

• 电弧研究与分析 •    下一篇

基于主成分分析和决策树的故障电弧识别方法

林靖怡1, 武建文1, 李奎2, 邵阳1, 王尧2, 马明顺1, 夏尚文1, 佟子昂1   

  1. 1.北京航空航天大学 自动化科学与电气工程学院, 北京 100191
    2.河北工业大学 电气工程学院, 天津 300131
  • 收稿日期:2023-02-04 出版日期:2023-04-30 发布日期:2023-05-09
  • 作者简介:林靖怡(1993—),女,博士研究生,研究方向为电器理论及应用。|武建文(1963—),男,教授,博士生导师,研究方向为开关电弧理论及应用、智能电器及电力电子技术。|李 奎(1965—),男,教授,博士生导师,研究方向为开关电器可靠性与智能化。
  • 基金资助:
    国家自然科学基金项目(51937004)

Arc Fault Identification Method Based on Principal Component Analysis and Decision Tree

LIN Jingyi1, WU Jianwen1, LI Kui2, SHAO Yang1, WANG Yao2, MA Mingshun1, XIA Shangwen1, TONG Ziang1   

  1. 1. School of Automation Science and Electrical Engineering,Beihang University,Beijing 100191, China
    2. School of Electrical Engineering,Hebei University of Technology,Tianjin 300131, China
  • Received:2023-02-04 Online:2023-04-30 Published:2023-05-09

摘要:

针对故障电弧产生时故障电流受负载影响较大、不易识别的问题,提出基于主成分分析与决策树的故障电弧识别方法。采用主成分分析法对有故障电弧产生时电流的8种时域特征参数降维处理,将去除冗余信息后的3种主成分作为特征值输入决策树,建立分类模型。所提方法能有效简化决策树分类模型,减少故障电弧识别时间,故障电弧识别准确率可达95%。

关键词: 主成分分析, 时域特征, 决策树, 故障电弧识别

Abstract:

The fault current of the fault arc is affected by the load and is not easy to be identified.To solve this problem,the arc fault identification method based on principal component analysis and decision tree was proposed.The principal component analysis method was used to reduce the dimension ality of eight temporal characteristic parameters of current generated by the fault arc.After removing the redundant information,the three kinds of principal components were input into the decision tree as characteristic values to establish a classification model.The proposed method can effectively simplify the decision tree classification model,reduce the time of fault arc identification,and reach the accuracy of fault arc identification of 95%.

Key words: principal component analysis, time domain feature, decision tree, fault arc identification

中图分类号: