Rapid Identification Method for the Sudden Change of Operating Errors in Smart Electricity Meters

The sudden change of operating errors is a critical problem for the online monitoring of electricity meters at present. In this paper, we propose a fast identification method based on the matching of line loss residual distribution characteristic approximation, in an attempt to realize the error estimation and sudden change moment localization of electricity meters with sudden change in operating errors. First of all, based on Kirchhoff's law, a line loss mathematical model that accurately takes into account the power load information is constructed to achieve accurate quantification of line loss (LL), and the line loss coefficient and the operating error estimation of electricity meters during the training period are obtained through nonlinear constraint optimization; consequently, on the basis of the distribution characteristics of line loss residuals in the blind period, K-Nearest Neighbor (KNN) algorithm is used to select the corresponding reference period in the form of approximation matching; finally, in conformity to the difference in the characters of line loss residuals in the reference period and the blind period, a variety of statistical indicators are used to comprehensively judge the out-of-tolerance electricity meter and its out-of-tolerance range. The simulation analysis and the experimental verification results in the actual station confirm the effectiveness and feasibility of the algorithm.