济源职业技术学院学报2024,Vol.23Issue(2) :58-63.DOI:10.3969/j.issn.1672-0342.2024.02.011

基于随机森林法的GIS内隔离开关机械故障诊断研究

Researches on Mechanical Fault Diagnosis of GIS Internal Isolation Switch Based on Random Forest

代书海 田瑞 曹双鹏
济源职业技术学院学报2024,Vol.23Issue(2) :58-63.DOI:10.3969/j.issn.1672-0342.2024.02.011

基于随机森林法的GIS内隔离开关机械故障诊断研究

Researches on Mechanical Fault Diagnosis of GIS Internal Isolation Switch Based on Random Forest

代书海 1田瑞 2曹双鹏3
扫码查看

作者信息

  • 1. 吉林电力股份有限公司松花江热电有限公司,吉林吉林 132000
  • 2. 国网辽宁省电力有限公司超高压分公司,辽宁沈阳 110000
  • 3. 国网辽宁省电力有限公司辽阳供电公司,辽宁辽阳 111000
  • 折叠

摘要

提出了基于驱动电机绕组电流与随机森林法的GIS内隔离开关机械故障诊断方法.将隔离开关驱动电机在不同操作阶段的极大电流值及其出现时刻作为特征值,通过力矩加载工装对隔离开关故障率较高的三类四种机械故障进行了模拟实验,并建立基于随机森林法的隔离开关机械故障诊断模型.最后,基于上述分类模型对模拟故障类型进行分类.分类结果表明,基于随机森林的隔离开关机械故障诊断的准确率高达90%,可有效实现隔离开关机械故障的诊断.

Abstract

A mechanical fault diagnosis method of GIS isolation switch based on the winding current of drive motor and random forest is proposed.The maximum current value of the motor driven by the isolation switch in different operating stages and its appearance time are taken as the characteristic value.Secondly,three types and four kinds of mechanical faults with high failure rate of isolation switch are simulated by using torque loading tool,and a mechanical fault diagnosis model of isolation switch based on random forest is established.Finally,based on the above classification model,the simulated fault types are analyzed,and the classification results show that the mechanical fault diagnosis accuracy of random forest based isolation switch is as high as 90%,which can effectively realize the diagnosis of isolation switch mechanical fault.

关键词

GIS内隔离开关/机械故障/随机森林/绕组电流/故障诊断

Key words

GIS internal disconnector/mechanical failure/random forests/winding current/fault diagnosis

引用本文复制引用

出版年

2024
济源职业技术学院学报
济源职业技术学院

济源职业技术学院学报

影响因子:0.274
ISSN:1672-0342
段落导航相关论文