南方电网技术2024，Vol.18Issue(12)：51-61,11.DOI:10.13648/j.cnki.issn1674-0629.2024.12.007
基于RFC-SAGA-RBF的直流偏磁下CT畸变电流反衍方法
Inverse Propagation Method of CT Distortion Current Under DC Bias Based on RFC-SAGA-RBF
摘要
Abstract
The DC bias current in the transmission line can easily cause the saturation of current transformer(CT)and the harmonic components contained in the secondary current output of CT in saturation state will lead to an increase in power energy metering error.In order to improve the accuracy of current transformer measurement under DC bias,an inverse propagation method is proposed for CT distortion current.In this method,the inverse propagation process of CT distortion current under DC bias is divided into two stages:off-grid and on-grid.In the off-grid stage,the data sample set is generated by changing the operating environment of CT,and then the saturation degree of CT is classified by random forest classification(RFC)algorithm.Finally,a simulate anneal genetic algorithm-radial basis function(SAGA-RBF)model is trained for each subclass to simulate the saturation current.In the on-grid stage,the saturation data segment of the secondary current waveform is extracted by wavelet transform,and then the saturation data segment is input into the off-line model to realize the inverse propagation of the secondary distorted current.The simulation results show that the proposed method can relike the inverse propagation of the primary side current,and improve the measurement ac-curacy of CT.关键词
电磁式电流互感器/直流偏磁/饱和电流重构/小波变换/随机森林分类算法Key words
electromagnetic current transformer/DC bias/saturation current reconstruction/wavelet transform/RFC algorithm分类
信息技术与安全科学引用本文复制引用
王波,尹仕红,肖勇,胡珊珊,范小飞,潘深琛,王保帅..基于RFC-SAGA-RBF的直流偏磁下CT畸变电流反衍方法[J].南方电网技术,2024,18(12):51-61,11.基金项目
广东省科技计划资助项目(2021B1212050014) （2021B1212050014）
中国南方电网有限责任公司科技项目(090000KK52220012).Supported by the Science and Technology Plan Project of Guangdong Province(2021B1212050014) （090000KK52220012）
the Science and Technology Project of China Southern Power Grid Co.,Ltd.(090000KK52220012). （090000KK52220012）
