Estimation Method of Line Loss Rate in Low Voltage Area Based on Mean Shift Clustering and BP Neural Network

Abstract The main problems faced in the line loss management of the distribution network are the incomplete meter configuration, the difficulty of collecting operating data, and the excessive number of components and nodes. These problems lead to a very complicated calculation of line loss rate. This paper proposes an improved BP neural network estimation method for passive low voltage area line loss rate driven by low voltage area characteristic data, and realizes it through programming. First, the characteristic data required for calculating the line loss rate of the low-voltage passive station area is selected and classified according to the station area capacity after standardization. The BP neural network model improved by Mean Shift clustering method is used to calculate the line loss rate for the station area data of the same capacity, which provides scientific basis and data support for the calculation of the station area line loss rate and the management of high-loss stations in the future. A passive low-voltage station with a capacity of 315KVA in Zhejiang Province was used as the modeling object to perform simulation calculations to verify the accuracy of the proposed method.