Title :
Prediction of dynamic hysteresis losses under nonsinusoidal exciting field by using neural networks and harmonic analysis
Author :
Salvini, A. ; Coltelli, C.
Author_Institution :
Dipt. di Ingegneria Elettronica, Rome Univ., Italy
Abstract :
Summary form only given. In this paper a simple approach to predict the power losses into a ferromagnetic nucleus when it is excited by a distorted field, H(t) is presented. The aim is also to investigate on the contribute to the losses of each harmonic of the imposed magnetic field. The first step of the method is the use of Back propagation Neural Network (NN) and the actual frequency transplantation (AFT) to predict the loop shape, as described by Salvini et al. (IEEE Trans. Magn. vol37, p.3315-3319, (2001)). As a successive step, the Fourier Descriptor (FD) is applied to the sampled point of the predicted loop on the B-H plane to evaluate the Fourier Series (FS) of the flux density B(t). This last evaluation allows one to calculate the power losses.
Keywords :
Fourier series; eddy current losses; harmonic analysis; magnetic flux; magnetic hysteresis; neural nets; physics computing; Back propagation Neural Network; dynamic hysteresis losses; ferromagnetic nucleus; flux density; frequency transplantation; harmonic analysis; neural networks; nonsinusoidal exciting field; Frequency; Gyroscopes; Harmonic analysis; Hysteresis motors; Magnetic hysteresis; Neural networks; Power system harmonics; Synchronous motors; Toroidal magnetic fields; Torque;
Conference_Titel :
Magnetics Conference, 2002. INTERMAG Europe 2002. Digest of Technical Papers. 2002 IEEE International
Conference_Location :
Amsterdam, The Netherlands
Print_ISBN :
0-7803-7365-0
DOI :
10.1109/INTMAG.2002.1001352