CNN在输配电网络故障诊断中的应用  

Application of CNN in Fault Diagnosis of Transmission and Distribution Networks

作　　者：姜淑娅 JIANG Shuya(Shandong Jigang Zhongdian Intelligent Technology Co.,Ltd.,Jinan 250000,China)

机构地区：[1]山东济钢众电智能科技有限公司,山东济南250000

出　　处：《通信电源技术》2025年第5期225-227,共3页Telecom Power Technology

摘　　要：为提升输配电网络运行质量,提高故障识别的智能化水平,以卷积神经网络(Convolutional Neural Network,CNN)为例进行研究。通过构建基于CNN的输配电网络故障识别模型进行故障定位与分类,借助仿真试验对比CNN模型与支持向量机(SupportVectorMachine,SVM)识别模型、决策树识别模型,验证所提模型的效果。结果表明,CNN方法在精度、召回率、实时性等方面均优于其他两种方法,能够有效提高针对输配电网络的故障识别准确性与效率,为电力系统的智能化运维提供重要参考。In order to improve the operation quality of transmission and distribution network and improve the intelligent level of fault identification,this paper takes the Convolutional Neural Network(CNN)as an example to study.Fault location and classification are carried out by constructing a fault identification model of transmission and distribution network based on CNN,and the effect of the proposed model is verified by comparing CNN model with Support Vector Machine(SVM)identification model and decision tree identification model through simulation experiments.The results show that CNN method is superior to the other two methods in accuracy,recall and real-time,which can effectively improve the accuracy and efficiency of fault identification for transmission and distribution networks and provide important reference for intelligent operation and maintenance of power systems.

关 键 词：输配电 网络故障 卷积神经网络(CNN) 

分 类 号：TM7[电气工程—电力系统及自动化]