Electronic converter models implemented with radial basis function networks
María Ángeles Moreno, Julio Usaola · 2004
Radial basis function (RBF) neural networks can be applied to the modelling of electronic converters. In this paper a new steady-state model of an uncontrolled bridge rectifier with capacitive DC smoothing is presented. The model considers the commutation effect and allows to obtain the harmonic currents injected by the converter in magnitude and angle. The model can be used to evaluate the harmonic distortion introduced in balanced networks by this device. The technique can be applied to any other converter or any other nonlinear load. Hence it is no more needed to know the analytical relationship between harmonic voltages and currents.