An approach based on artificial neural networks (ANNs) for modeling the average behavior of nonlinear power electronic converters is explored in this paper. The aim herein is to analyze the effectiveness of nonlinear autoregressive exogenous (NARX) ANN to realize black-box average dynamic models of dc-dc converters, capturing nonlinearities and being valid both in time and frequency domains. The proposed approach is evaluated by means of simulation and experimental results on a boost converter, taken as a test case. Verification is performed by comparing, at different operating points, the time responses and transfer functions provided by the ANN-based modeling, particularly considering the nonlinearity constituted by the discontinuous conduction mode (DCM) and the continuous conduction mode (CCM), to be correctly represented by the same model.
Modeling Non-Linearities of Power Electronic Converters Using Artificial Neural Networks
Zilio A.;Biadene D.;Caldognetto T.;Mattavelli P.
2023
Abstract
An approach based on artificial neural networks (ANNs) for modeling the average behavior of nonlinear power electronic converters is explored in this paper. The aim herein is to analyze the effectiveness of nonlinear autoregressive exogenous (NARX) ANN to realize black-box average dynamic models of dc-dc converters, capturing nonlinearities and being valid both in time and frequency domains. The proposed approach is evaluated by means of simulation and experimental results on a boost converter, taken as a test case. Verification is performed by comparing, at different operating points, the time responses and transfer functions provided by the ANN-based modeling, particularly considering the nonlinearity constituted by the discontinuous conduction mode (DCM) and the continuous conduction mode (CCM), to be correctly represented by the same model.Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.