Reconstruction of Missing Measurement Data in Distribution Networks Based on Signal Features and Naive Bayesian Classifiers

This paper proposes a method for reconstructing missing measurement data in distribution networks based on signal features and naive Bayes classifiers. By analyzing the time-frequency characteristics of the transient current missing data signal measured by the power meter in the distribution network, the time-frequency characteristic parameters of the signal are obtained, and the characteristic distribution and correlation of the signal in the time and frequency domains are comprehensively analyzed. The Naive Bayes classifier is used to extract and divide the time-frequency characteristics of the missing measurement data signal of the distribution network. Based on this feature, random orthogonal matrices, coefficient matrices, and singular values are added to the reconstruction model to complete the reconstruction of missing measurement data in the distribution network. Experimental results have shown that this method can not only reduce the time consumption in the reconstruction process of distribution network measurement data, but also quickly and accurately reconstruct missing reactive power data without affecting existing data.