首页 | 官方网站   微博 | 高级检索  
     

基于时间序列压缩动态时间弯曲距离故障区段定位
引用本文:许冲冲,罗勋华,郭上华,辛锋,王志勇,杨志祥.基于时间序列压缩动态时间弯曲距离故障区段定位[J].电测与仪表,2019,56(10):127-133.
作者姓名:许冲冲  罗勋华  郭上华  辛锋  王志勇  杨志祥
作者单位:国网北京市电力公司,北京,100045;珠海许继电气有限公司,广东珠海,519070
基金项目:国家电网有限公司总部科技项目;广东省科技计划
摘    要:针对基于暂态零模电流相关系数法的配电网故障区段定位方法存在数据处理量大且需要时间同步问题,提出一种基于时间序列压缩的动态时间弯曲(DTW)距离的小电流接地故障区段定位方法。先对终端控制单元(FTU)采集的故障录波数据预先处理,只提取故障发生后一个工频周波内的初始值、极值以及两极值点之间数据变化差值最大的两点,再将新的时间序列上传至主站,最后由主站采用动态时间弯曲距离算法求取两相邻FTU零模电流的相似性来进行故障区段定位。通过算法对比分析可得,在抗同步误差能力和信号幅值反应能力上,DTW算法较强于相关系数法;在直流分量表达能力上,两算法效果基本一致。此外,数据压缩后的新时间序列在DTW距离上的相似度表达能力基本与原时间序列相同,但数据处理量减少了至少一半以上。大量ATP/EMTP仿真说明文中所提方法能在不同故障角、接地电阻、故障位置下准确判出故障区段。

关 键 词:配电网  故障区段定位  数据压缩  动态时间弯曲距离
收稿时间:2018/4/20 0:00:00
修稿时间:2018/4/20 0:00:00

A%20Fault%20Section%20Location%20Method%20Based%20on%20Dynamic%20Time%20Warping%20Distance%20with%20Compress%20Time%20Sequence
XU Chongchong,LUO Xunhu,GUO Shanghu,XIN Feng,WANG Zhiyong and YANG Zhixiang.A%20Fault%20Section%20Location%20Method%20Based%20on%20Dynamic%20Time%20Warping%20Distance%20with%20Compress%20Time%20Sequence[J].Electrical Measurement & Instrumentation,2019,56(10):127-133.
Authors:XU Chongchong  LUO Xunhu  GUO Shanghu  XIN Feng  WANG Zhiyong and YANG Zhixiang
Affiliation:State Grid Beijing Electric Power Corporation,Zhuhai XJ Electric Cor Ltd,Zhuhai XJ Electric Cor Ltd,State Grid Beijing Electric Power Corporation,State Grid Beijing Electric Power Corporation,Zhuhai XJ Electric Cor Ltd
Abstract:Since%20the%20method%20of%20transient%200%20mode%20current%20correlation%20coefficient%20in%20distribution%20network%20fault%20section%20locating%20exists%20the%20problem%20with%20large%20amount%20data%20processing%20and%20need%20time%20synchronization,%20this%20paper%20presents%20a%20dynamic%20time%20warping%20(DTW)%20distance%20method%20based%20on%20time%20series%20compression%20for%20fault%20section%20locating.%20The%20proposed%20method%20only%20extract%20data%20within%20a%20power%20frequency%20cycle%20after%20fault%20occurs%20which%20include%20the%20initial%20value,%20extreme%20value%20and%20two%20points%20of%20maximum%20difference%20change%20between%20two%20extreme%20points.%20Then%20a%20new%20time%20sequence%20is%20uploaded to the%20host,%20and%20finally%20by%20the%20main%20dynamic%20time%20warping%20distance%20algorithm%20calculating%20the%20similarity%20of%20two%20adjacent%20FTU.%20Through%20the%20comparison%20of%20the%20algorithm,%20the%20DTW%20algorithm%20is%20better%20than%20the%20correlation%20coefficient%20method%20in%20the%20ability%20of%20anti-synchronization%20error%20and%20signal%20amplitude.%20In%20addition,%20the%20similarity%20calculation%20of%20the%20new%20time%20series%20on%20DTW%20distance%20is%20the%20same%20as%20the%20original%20time%20series,%20but%20the%20data%20processing%20capacity%20is%20reduced%20by%20more%20than%20half.%20A%20large%20amount%20of%20ATP/EMTP%20simulation%20shows%20that%20the%20proposed%20method%20can%20accurately%20find%20fault%20segments%20in%20different%20fault%20angles,%20grounding%20resistance%20and%20fault%20location.
Keywords:distribution%20network  %20fault%20section%20location  %20data%20compression  %20dynamic%20time%20warping%20distance
本文献已被 CNKI 万方数据 等数据库收录!
点击此处可从《电测与仪表》浏览原始摘要信息
点击此处可从《电测与仪表》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司    京ICP备09084417号-23

京公网安备 11010802026262号