单端行波故障测距的组合方法研究  

A Combined Method for Single-Ended Traveling Wave Fault Location

作　　者：梁睿[1] 孙式想[1] 

机构地区：[1]中国矿业大学信息与电气工程学院,江苏省徐州市221116

出　　处：《电网技术》2013年第3期699-706,共8页Power System Technology

基　　金：国家自然科学基金资助项目(51107143);中央高校基本科研业务费专项资金资助(2011QNA19)~~

摘　　要：单端辐射状电网发生单相接地故障时,单端行波故障测距的关键是正确识别来自故障点的反射波、对端母线的反射波及健全线路的反射波。行波零模波速具有较强的依频特性,而线模波速较稳定。由于零模波速的不稳定性,当取其为固定值时,导致基于线模、零模波速差的单端行波故障测距精确度不够。针对此问题,首先利用神经网络训练学习,估算出零模波速,基于线模、零模波速差得到初步故障距离,以此初步故障距离从线模行波突变波头中找出故障点反射波头及对端母线反射波头,再把线模波速代入相应的测距公式得到精确故障距离。利用PSCAD/EMTDC搭建仿真模型,对不同的故障情况进行仿真,均验证了该方法具有较高的测量精度和可靠性。When single-phase-to-earth fault occurred in radial distribution network supplied by single power,the key problem of the single-ended traveling wave fault location lies in the correct recognition of the reflected waves coming from fault point,bus at opposite terminal of the faulty line and healthy feeders respectively.The zero modulus velocity of traveling wave is evidently related to the frequency and the aerial mode velocity of traveling wave is more stable.The instability of the zero modulus velocity leads to insufficient fault location accuracy from single-ended traveling wave fault location based on wave velocity contrast between zero modulus velocity and aerial mode velocity when a constant value is taken for zero modulus velocity.In allusion to this problem,firstly through the training and learning by neural network,the zero modulus velocity is estimated and based on the wave velocity contrast between zero modulus velocity and aerial mode velocity a preliminary fault distance is attained;secondly,using the preliminary fault distance,the reflected wave fronts from fault point and the bus at opposite terminal of the faulty line are found out from the sudden change wavefront of aerial mode traveling wave;thirdly,an accurate fault distance can be achieved by substituting aerial mode velocity into corresponding fault location formula.Using a simulation model built by PSCAD/EMTDC to simulate different fault conditions,results of simulation show that the fault location by the proposed method is reliable and accurate.

关 键 词：波速差 小波变换 神经网络 波头识别 故障测距 

分 类 号：TM771[电气工程—电力系统及自动化]