作者: A. Salvini , C. Coltelli
DOI: 10.1109/20.952603
关键词:
摘要: Neural Network (NN) and actual frequency transplantation (AFT) are combined for prediction of dynamic hysteresis when the exciting field, H(t), is highly polluted by harmonics. The NN forecasts Fourier Series flux density well-known H(t) waveforms (i.e., triangular, square wave fields etc.). task AFT to approach arbitrary distortion exploiting loop predictions under pure sinusoidal excitations then transplanting branches related frequencies detected in short time-windows period. These will be evaluated an appropriate time-frequency analysis H(t). Model validations presented comparison with experimental data.