基于深度学习的电力系统谐波电能计量误差校正方法  

Error Correction Method for Harmonic Energy Measurement of Power System Based on Deep Learning

作　　者：梁丽 马亚珍 LIANG Li;MA Ya-zhen(Gansu Electric Power Corporation Marketing Division Metering Center of SGCC,Lanzhou 730300 China)

机构地区：[1]国网甘肃省电力公司市场营销事业部计量中心,甘肃兰州730300

出　　处：《自动化技术与应用》2023年第9期49-52,57,共5页Techniques of Automation and Applications

摘　　要：传统电力系统电能计量中由于未及时处理谐波电流影响,导致无法有效检测出电力系统的比差、角差拟合曲线,谐波振幅数据,导致检测误差大的问题。提出基于深度学习的电力系统谐波电能计量误差校正方法。通过深度学习分析影响谐波电流的因素;再利用动态负荷算法计算电力系统的电流直流偏置比差与角差;最后将计算结果整合,实现电力系统谐波电能计量的误差校正。实验结果表明,利用所提方法误差校正时,能有效检测电力系统的比差拟合曲线、角差拟合曲线,校正时谐波振幅数据检测误差较小。In the traditional power system energy measurement,the influence of harmonic current is not dealt with in time,which leads to the problem that the ratio difference,angle difference fitting curve and harmonic amplitude data of the power system can not be detected effectively,which leads to large detection error.This paper proposes a power system harmonic energy measurement error correction method based on deep learning.The factors influencing harmonic current are analyzed by deep learning;Then,the dynamic load algorithm is used to calculate the DC bias ratio and angle difference;Finally,the calculation results are integrated to realize the error correction of harmonic energy measurement in power system.The experimental results show that the proposed method can effectively detect the ratio difference fitting curve and angle difference fitting curve of power system,and the detection error of harmonic amplitude data is small.

关 键 词：深度学习 电力系统 谐波电能计量 误差校正 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]