An Evaluation Method of 10k V Distribution Network Line Loss Based on Improved BP Neural Network

Abstract

A novel method of calculating 10kV distribution network line loss is proposed and realized by programming, which is improved BP neural network model based on adaptive genetic algorithm. Firstly, the characteristic index system is established according to electric characteristic parameters of samples. Then, through leaning the training samples by improved model of BPNN, the line loss evaluation model is obtained. After that,10kV line loss of test samples can be evaluated actually. The improved algorithm of BPNN is adopted to fit complex nonlinear relation between line loss and electric characteristic parameters. The 10kV distribution network in a real system is taken as an example. The accuracy of the proposed method is verified by simulation and calculation of the example. Compared with traditional BPNN, This method has the advantages of fast convergence and high accuracy.