Prediction of harmonic currents in residential load windows, by artificial neuronal models

Abstract

At present, most of the residential loads connected to the electric power distribution systems are non-linear, ie the current contains harmonic components as well as the voltage loop. These devices inject harmonics into the mains supply at the common connection point. For the analysis of the nonlinearity present in these loads, it is necessary to predict the harmonic currents generated by this type of equipment or load groups. In this paper, models of the harmonic behavior of nonlinear residential electrical equipment are presented using artificial redneurons. The analyzed equipment has as main characteristic that its electrical power supply system is based on single-phase full-wave rectifiers with capacitive filter