An Identification Method Of Load Harmonic Current Based On BP Neural Network

As the harmonic pollution is getting m ore and more serious, there is often some dispute between the electric supp ly com panies and cus tomers. In order to m ake the identification of harmonic current produced by the lo ad, the paper pres ents an identification method based on the BP neur al network, thus m aking an objective and reasonable assessment method of the load harm onic emission level. Based on the theory of dif ferential equation m odel of nonlinea r load, an identifica tion m ethod of load harmonic current using BP neural networ k is proposed. Considering tim e-varying frequency, the fundamental frequency and voltage on the load can be determ ined by the windowed discrete Fourier transform and double spectral line interpolation. In order to improve the generalization ability of BP neural network, voltage and current data measured at the connection point of utility grid is checked and Bayesian regularization algorithm is adopted. W ith the trained BP neural network describing the nonlinear load, the current incented by the funda mental voltage can be obtained. The simulation results demonstrate that the total harm onic distortion of the load current based on BP neural network is almost independent of power cap acity and harmonic voltage within the range of utility gr id harmonic voltage lim its, which is beneficial to the division of harmonic responsibility and harmonic control.