基于PCA-SVR的智能电表计量误差检测方法设计  

Design of Measurement Error Detection Method of Intelligent Electric Meter Based on PCA-SVR

作　　者：宋强 吴才远 杨婧 SONG Qiang;WU Caiyuan;YANG Jing(Measurement Center,Guizhou Power Grid Co.,Ltd.,Guiyang 550000,China)

机构地区：[1]贵州电网有限责任公司,计量中心,贵州贵阳550000

出　　处：《微型电脑应用》2025年第2期98-101,共4页Microcomputer Applications

基　　金：贵州电网有限责任公司营销技改项目(0600002GY62210006)。

摘　　要：为减少智能电表计量误差,提高检测准确性,对基于主成分分析法—回归型支持向量机(PCA-SVR)的智能电表计量误差检测方法进行设计研究。利用PCA模型分析智能电表互感器信号有效值,以Q统计量作为指标,检测智能电表的计量状态。在正常计量状态下构建智能电表计量误差降维模型,利用SVR在实际工作情况下智能电表的计量误差。使用模拟实验平台开展智能电表计量模拟实验,结果显示,设计方法能够完成智能电表计量误差状态评价,并且添加电网故障干扰后,仍旧能够准确检测出智能电表的计量误差,且检测结果与提前设置结果较为接近,说明该方法检测智能电表计量误差时准确性较高,具有一定应用价值。In order to reduce the measurement error of smart meters and improve the detection accuracy,a measurement error detection method of smart meters based on principal component analysis-regression support vector machine(PCA-SVR)is designed and studied.The principal component analysis(PCA)model is used to analyze the effective value of the transformer signal of the smart meter,and the Q statistic is used as an indicator to detect the measurement tate of the smart meter.The dimension reduction model of the measurement error of the smart meter is built under the normal metering state,and the regression support vector machine(SVR)is used to calculate the measurement error of the smart meter under the actual working condition.This paper uses a simulation test platform to carry out the measurement simulation test of smart meters,the results show that the design method can complete the state evaluation of measurement errors of the smart meter,and can still accurately detect the measurement error of smart meters after adding the power grid fault interference.The detection result is close to the previous setting result,which indicates that the method has higher accuracy in detecting the measurement error of smart meters,and has certain application value.

关 键 词：主成分分析 回归型支持向量机 信号有效值 智能电表 计量误差 误差降维 

分 类 号：TH-3[机械工程]