基于Elman神经网络的煤矿井下电网故障选线研究  

Research on Faulty Line Detection of Mine Underground Electric Power Network Based on Elman Neural Network

作　　者：李陈陈[1] 李东辰[1] 皇淼淼[1] 许威[1] 

机构地区：[1]安徽理工大学电气与信息工程学院,安徽淮南232001

出　　处：《煤矿机械》2012年第10期290-292,共3页Coal Mine Machinery

摘　　要：为了解决煤矿井下电网发生单相接地故障时的选线问题,将Elman神经网络引入故障选线领域,根据故障与非故障线路零序电流中谐波电流含量不同的特征,选出故障线路。采用零序电流中的三次、五次、七次谐波电流含量构成特征相量作为Elman神经网络的输入向量对网络进行训练,并用训练好的网络选取故障线路。为了验证该方法的正确性和有效性,在simulink平台上搭建煤矿井下电网仿真系统。In order to solve problem of faulty line detection when one-phase ground fault happens in mine underground electric power network,Elman neural network is introduced into faulty line detection field.The faulty line is picked out according to characteristics of harmonic current content in zero sequence current of fault and non-fault lines.The feature vector is formed by three,five,seven times harmonic current content of zero sequence current,and as elman neural network＇s input vector to train network.And the faulty line is selected out through trained network.In order to validate method is correct and effective,the simulation system of mine underground electric power network is built in simulink platform.

关 键 词：ELMAN神经网络 故障选线 零序电流 谐波电流含量 

分 类 号：TD611[矿业工程—矿山机电]