摘要
针对当前智能电表现场检定效率低、人力成本高、实时性差、无法全量监测等问题,提出一种基于动态线损和渐消记忆递推最小二乘法(dynamic line loss and fading memory recursive least square,DLL-FMRLS)的智能电表误差在线估计算法。首先,通过分析台区线损与供电量之间的关系,对传统模型进行改进,提出动态线损误差模型,该模型中线损可随实际供电量变化,使得模型获得的误差估计值更接近实际值;然后,利用FMRLS算法求解动态线损误差模型,以获得智能电表运行误差;最后,根据某省电网公司的实际数据对算法现场验证。结果结果表明,与列文伯格–马夸尔特(Levenberg-Marquardt, LM)算法和限定记忆最小二乘(limited memory recursive least squares,LMRLS)算法相比,所提算法可以有效提高智能电表的误差估计的准确度。
The current smart meter verification methods have problems such as low verification efficiency, high labor cost, poor real-time performance and difficulty in full coverage.To solve the above problems, this paper proposed a smart meter error online estimation algorithm based on dynamic line loss and fading memory recursive least square method(DLL-FMRLS). Firstly, through the analysis of the relationship between the line loss in the station area and the power supply,the dynamic line loss error model was proposed by improving the traditional model. In this model, the estimated error obtained was closer to the actual value because the line loss could vary with the actual power supply. Secondly, the FMRLS algorithm was used to solve the dynamic line loss error model to obtain the operating error of the smart meter. Finally, the algorithm was verified according to the actual data provided by a provincial power grid company. Compared with the Levenberg-Marquarelt(LM) and limited memory recursive least squares(LMRLS) algorithms, the proposed algorithm are superior in the accuracy of the error estimation of the smart meter.
作者
徐焕增
孔政敏
王帅
苏志华
XU Huanzeng;KONG Zhengmin;WANG Shuai;SU Zhihua(School of Electrical Engineering and Automation(Wuhan University),Wuhan 430072,Hubei Province,China)
出处
《中国电机工程学报》
EI
CSCD
北大核心
2021年第24期8349-8357,共9页
Proceedings of the CSEE
关键词
智能电表
误差估计
动态线损
渐消记忆递推最小二乘法
在线估计
smart meter
error estimation
dynamic line loss
fading memory recursive least square method
online estimation
