基于C型行波与SVM的配电线路故障定位  

Composite Fault Location Method Based on C-traveling Wave and SVM for Distribution Lines

作　　者：严凤[1] 李双双[1] 

机构地区：[1]华北电力大学电气与电子工程学院,保定071003

出　　处：《电力系统及其自动化学报》2016年第1期86-90,共5页Proceedings of the CSU-EPSA

摘　　要：在进一步研究C型行波故障定位方法的基础上,通过对现有配电网故障定位方法的分析,针对我国配电线路分支较多,故障信号难以捕捉的特点,提出了基于C型行波与支持向量机SVM(support vector machine)的配电线路综合故障定位的方法。该方法不仅融合了C型行波在故障测距中的优势与支持向量机在模式识别方面的功能,而且采用小波降噪技术对故障信号加以处理,通过优化SVM参数,并与BP人工神经网络法进行对比得出:SVM分类法能够准确判断故障区段。经过EMTP与Matlab仿真实验表明:C型行波与支持向量机相互结合的方法能够准确地对带有多分支的配电网进行单相接地故障的定位。A comprehensive C-type of traveling wave-support vector machine （SVM） fault location method is presented after analyzing the existing fault location methods and the characteristics of the distribution network. The lines in our distribution network have lots of branches and the fault signal is difficult to be captured. By using the advantages of C-traveling wave fault location method in distance measuring and the pattern recognition function of support vector machines, the method can improve the accuracy of fauh location. It also adopts the wavelet to reduce the noise of fault signal, and the accuracy can be more precise when the parameters of SVM are optimized. Comparing with the BP artificial neural network, SVM is more effective to find the fault section. The simulation results of EMTP and Matlab show that the method combined C-type of traveling wave and support vector machine can determine the single phase-to-earth fault point in distribution networks with branches accurately.

关 键 词：配电网 故障定位 C型行波 模式识别 支持向量机 

分 类 号：TM711[电气工程—电力系统及自动化]