基于CNN-SVM算法的高压输电线路故障识别研究  

Research on High Voltage Transmission Line Fault Identification Based on CNN-SVM Algorithm

作　　者：宋文志 刘柯余 SONG Wenzhi;LIU Keyu(Beijing Guodiantong Network Technology Co.,Ltd.,Beijing 100192,China;Material Company of State Grid Sichuan Electric Power Company,Chengdu 610000,China)

机构地区：[1]北京国电通网络技术有限公司,北京100192 [2]国网四川省电力公司物资公司,四川成都610000

出　　处：《微型电脑应用》2025年第2期284-287,共4页Microcomputer Applications

摘　　要：在输电线路中出现故障将会对电网运行安全造成不利影响。为了提高高压输电线路稳定性,设计一种基于卷积神经网络-支持向量机(CNN-SVM)算法的高压输电线路故障诊断方法。构建SVM模型来判断相间故障接地状态,实现输电线路故障精确诊断。测试结果表明,单相和三相接地故障都实现了很高的识别精度,而只考虑初步识别过程对相间故障和相间接地故障的识别率都很低。采用本文方法可以获得更高的准确率,能够实现的相间故障更高精度识别效果,识别AB相间故障时达到了99.87%的准确率。When the transmission line fails in use,the safety of the power grid is adversely affected.In order to improve the stability of HV transmission lines,HV transmission line a fault diagnosis method based on convolutional neural networks-support vector machine(CNN-SVM)algorithm is designed.The SVM model is constructed to judge the phase fault grounding state,and the transmission line fault diagnosis is realized accurately.The test results show that the identification accuracy of both single-phase and three-phase ground faults is very high,but the initial identification process only achieves a very low recognition rate of both phase and phase ground faults.The method presented in this paper can obtain higher accuracy and achieve higher precision identification effect of phase-to-phase faults.The accuracy of the proposed method is 99.87%in identifying AB phase faults.

关 键 词：输电线路 故障识别 CNN SVM 特征提取 故障接地 

分 类 号：TP27[自动化与计算机技术—检测技术与自动化装置]