基于暂态行波的铁路电力电缆故障定位系统设计与实现  

Design and Implementation of Railway Power Cable FaultLocation System Based on Transient Traveling Wave

作　　者：王小星 WANG Xiao-xing(Electrification Department Branch of China Railway First Survey And Design Institute Group Co.,Ltd.,Xi'an 710043 China)

机构地区：[1]中铁第一勘察设计院集团有限公司电化处,陕西西安710043

出　　处：《自动化技术与应用》2021年第5期28-31,共4页Techniques of Automation and Applications

摘　　要：本文基于人工神经网络(ANN)确定铁路电力电缆故障位置,采用基于阻抗的稳态信号或瞬态信号过程中特征提取的合适时间间隔方式。通过MATLAB对本文所设计模型进行了仿真,以传输线的单端获取的故障电流信号作为输入。该系统考虑了在不同运行条件下的不同类型的故障,提取了五个特征,分别是两个条件下最小到最大的增量、幅度、标准偏差、能量和信号平均值。结果通过ANN的均方误差进行评估,表明并非所有功能都适合在一次信号条件下提取。This paper is based on artificial neural network(ANN)to determine the fault location of railway power cables,and adopts the appropriate time interval method for feature extraction in the process of impedance-based steady-state signals or transient signals.The model designed in this paper is simulated by MATLAB,and the fault current signal obtained by the single-end of the transmission line is used as the input.The system considers different types of faults under different operating conditions,and extracts five features,namely the minimum to maximum increment,amplitude,standard deviation,energy and signal average under the two conditions.The results are evaluated by the mean square error of ANN,which indicates that not all functions are suitable for extraction under the condition of one signal.

关 键 词：人工神经网络 故障定位 特征提取 MATLAB 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]