基于EMD/HHT的变压器故障检测算法研究  

Research on transformer fault detection algorithm based on EMD/HHT

作　　者：张广怡 Zhang Guangyi(Zhengzhou Electric Power College,Zhengzhou 450000,China)

机构地区：[1]郑州电力高等专科学校,河南郑州450000

出　　处：《无线互联科技》2023年第20期145-147,共3页Wireless Internet Technology

基　　金：河南省高等职业院校青年骨干教师项目,项目名称:基于EMD/HTT的变压器故障监测神经网络算法的研究,项目编号:2020GZGG058。

摘　　要：为精准实现变压器故障检测,文章提出基于EMD/HHT的变压器故障检测算法。该算法依据EMD算法对采集的变压器运行信号进行经验模态分解,获取多个固有模态分量;利用HHT算法分析固有模态分量的时频变化情况,获取其频率谱并提取变压器运行信号特征频率,将提取结果输入卷积神经网络模型中,通过网络模型的学习和训练,得出变压器故障检测结果。测试结果显示:该方法具有较好的应用性能,能够有效获取变压器的运行信号固有模态分量,通过所提取分量的频率特征精准完成不同类别变压器故障检测,为电力系统稳定运行提供了可靠的依据。In order to realize transformer fault detection accurately,a transformer fault detection algorithm based on EMD/HHT is proposed.This algorithm performs empirical mode decomposition on the collected transformer operation signals based on the EMD algorithm to obtain multiple intrinsic mode components.It uses the HHT algorithm to analyze the time-frequency changes of the inherent modal components,obtain their frequency spectrum,extract the characteristic frequency of the transformer operation signal,input the extracted results into the convolutional neural network model,and output the transformer fault detection results through learning and training of the network model.The test results show that this method has good application performance and can effectively obtain the inherent mode components of the transformer operation signal,reliably extract the frequency characteristics of the components,accurately detect different types of transformer faults,and provide a reliable basis for the stable operation of the power system.

关 键 词：EMD HHT 变压器 故障检测 特征频率 模态分量 

分 类 号：TM721[电气工程—电力系统及自动化]