Error Analysis Model of Electricity Meters Based on Recursive Weighted Least Squares

LIAO He; LIU Fei; YAN Yonghui; WU Jia; YU Wei

您当前的位置：

首页 >

文章列表页 >

Error Analysis Model of Electricity Meters Based on Recursive Weighted Least Squares

更新时间：2026-04-08

- Error Analysis Model of Electricity Meters Based on Recursive Weighted Least Squares
- Issue 2, (2026)
- 作者机构：
  
  江苏方天电力技术有限公司
- 作者简介：
- 基金信息：
- DOI：
  CLC：
- Published：2026
- 稿件说明：
移动端阅览
LIAO He, LIU Fei, YAN Yonghui, et al. Error Analysis Model of Electricity Meters Based on Recursive Weighted Least Squares[J]. 2026, (2).
DOI：

LIAO He, LIU Fei, YAN Yonghui, et al. Error Analysis Model of Electricity Meters Based on Recursive Weighted Least Squares[J]. 2026, (2). DOI：

摘要

为精准估计电能表运行计量误差，本文构建基于一种基于加权递推最小二乘法的电能表运行误差分析模型。首先，基于标定电流法识别轻载用户表，通过统计其在时间窗口内的轻载占比率，将多个轻载用户表合并为一个“虚拟用户表”，剔除误差偏高的轻载量测数据，同时保留整体数据完整性；其次，构建压降法线损模型，通过台区总表与各用户表的电压、电流数据直接计算线路压降引起的可变线损；然后，根据能量守恒定律，建立用户表运行误差分析线性方程；最后，基于加权递推最小二乘法求解用户表误差率，引入可变遗忘因子，动态调节历史数据权重，估算电能表分表误差。实验结果表明，相较于固定线损模型和动态线损模型，本模型可有效提高电能表误差估算的精确度，对已知超差表的检出率达100%。

Abstract

To accurately estimate the metering error of electricity meters

a running error analysis model based on weighted recursive least squares is proposed. Firstly

the light-load user meters are identified based on the calibration current method. By sta-tistically analyzing the light-load proportion within the time window

multiple light-load user meters are combined into a "virtual user meter"

and the light-load measurement data with high error are eliminated while maintaining the integrity of the overall data. Secondly

a voltage drop line loss model is constructed to directly calculate the variable line loss caused by the line voltage drop through the volt-age and current data of the distribution transformer's total meter and each user meter. Then

based on the law of con-servation of energy

a linear equation for the error analysis of the user meter operation is established; finally

the weighted recursive least square method is used to solve the error rate of the user meter

and a variable forgetting factor is introduced to dynamically adjust the weight of historical data to estimate the error of the sub-meter of the electric energy meter. The experimental results show that compared with the fixed line loss model and the dynamic line loss model

this model can effectively improve the accuracy of the electricity meter error estimation

and the detection rate of the known out-of-tolerance meters reaches 100%.

关键词

Keywords

references

张程.电能计量误差与改进策略分析[J].集成电路应用, 2025,42(08):332-333.

胡涛,伍栋文,赵燕,等.智能电能表基本电参量测量超差原因分析[J].江西电力,2024,48(02):31-33+46.

国家电网有限公司.智能电能表功能规范:Q/GWD 10354—2020[S].北京:中国电力出版社,2021.

梦菲,郭舒洁.基于神经网络的智能电能表计量误差模型研究[J].电子元器件与信息技术,2023,7(12):144-147+152.

成达,李宁,熊素琴,等.基于改进型BP神经网络的智能电能表应力误差分析方法[J].自动化与仪器仪表,2022,(07):308-312．CHENG Da,LI Ning,XIONG Suqin,et al．Stress error analysis method of intelligent watt-hour meter based on improved BP neural network[J]. Automation Instrumentation,2022,(07):308-312 ．

陈叶,杨正宇,朱梦梦等. 基于递归神经网络的智能电表运行误差远程估计方法[J]．电子设计工程,2022, 30(23): 71-74,80.

杨婧,宋强,叶文波,等.基于多模型融合的低压用户电能表运行误差在线监测方法研究[J].电力大数据,2024,27(12):54-62.

吕玉玲,陈文礼,程瑛颖,等.基于约束优化模型的智能电表运行误差及日线损率联合估计方法[J].电网技术,2025,49(03):1257-1265.

贾轩,许吉凯,任艺婧,等.基于样本扩容和数据驱动的台区理论线损计算方法[J].山东大学学报(工学版),2025,55(03):158-164.

曹舒,张颖,魏晓莹,等.一种基于线损迭代的超差智能电能表筛查方法[J].电力需求侧管理,2022,24(04):111-116.

刘子悦,周强,陈佳佳,等.基于特征选择的IDBO-ELM配电网台区线损计算[J].电网与清洁能源,2025,41(05):48-57.

刘科研,贾东梨,王薇嘉,等.考虑负荷不平衡分布的低压配电网线损实用计算策略[J].电力需求侧管理,2021,23(06):81-86.

杨耿杰,韦先灿,高伟.基于改进动态线损估计法的超差智能电表识别[J].电网技术, 2022,46(09):3662-3671YANG Gengjie,WEI Xiancan,GAO Wei．Error Estimation Method for Smart Meters Considering Sudden Misalignment [J]．Power System Technology,2022,46(09):3662-3671 .

曹辉.电子式电能表的计量误差因素与应对策略分析[J].集成电路应用,2025,42(05):330-331.

穆举文.电能计量装置运行误差与状态评价模型研究[D].太原理工大学, 2022.

方磊,徐俊,赵逸,等.基于改进梯度下降的低压台区技术线损优化估计方法研究[J].电气传动,2025,55(06):64-71.

胡志强,张曼,程绪敏,等. 改进型电压损失法低压台区理论线损计算[J]. 电子设计工程,2023,31(13):123-127.

王力超,邵强.基于渐消记忆递推最小二乘算法的台区线损诊断技术研究[J].中国新技术新产品,2025,(22):64-66+83.

魏东升.智能算法在配电网结构优化中的应用研究[J].科技创新与应用,2025,15(36):177-180+184.

黄丹,周永刚,刘单华,等.基于AdaBoost集成学习算法的低压台区理论线损计算与分析[J].电测与仪表,2025,62(11):120-128.

安海云,周前,刘玉方,等.基于拓扑门控的数据-机理混合驱动的主动配电网线损计算方法[J].山东大学学报(工学版),2025,55(05):51-61.

林泽科,杨炀,沈苏红,等.基于台区日线损率曲线特征的线损分析治理[J].电工技术,2024,(03):96-99.

邱育胜.基于凸松弛的LASSO正则带约束稀疏优化模型的超差智能电表识别[J].电工技术,2025,(12):114-118+122.

余传祥,潘傲然,毛文鹏,等.基于ISCSO的智能电表误差和线损率联合评估模型[J].电力系统保护与控制,2025,53(13):117-127．YU Chuanxiang, PAN Aoran, MAO Wenpeng, et al．Joint evaluation model of smart meter error and line loss rate based on ISCSO [J]．Power System Protection and Control,2025, 53(13): 117-127．

虞亦隆,李佳鋆,王敏,等.基于VAR的电能表降维误差估计模型[J].电力大数据,2023,26(07):32-39.

王浩,杨鹏,李翀,等.联合改进SVR和渐消记忆递推最小二乘算法的电能表误差评估模型研究[J].电力科学与技术学报,2023,38(05):206-215.

周茂一,黄婷钰,刘子文,等.基于含可变遗忘因子递推最小二乘的电力系统惯量评估方法[J].电力工程技术,2025,44(05):148-158.

王淑云,李新家,严永辉,等.台区同期线损率标杆值与合理区间计算方法[J].电力大数据,2023,26(03):11-18.