Modified J-A model and parameter identification based on data mining

Data mining and soft computing techniques have been widely used in Jiles-Atherton (J-A) hysteresis model parameters identification for ferromagnetic materials or ferromagnetic composites. However, the model cannot be applied to magnetostrictive composites (MSC). That is because not only the nonmagnetic matrix will change the magnetic field distribution in the composites, but also the magnetostriction will be affected by the fabrication procedure. In order to realize the pre-estimation of the magnetostrictive composites magnetic properties, we present a new prediction method. This method is based on modified J-A hysteresis model, utilizing data mining technique to identify model parameters from the raw measured data of magnetostrictive alloy. A methodology including-experimental determination, MJA model, parameters identification by differential evaluation algorithm, is discussed in detail. Then, the experimental data of magnetostrictive composites are compared to the simulations. The theoretical model agrees with the measurement results very well. Our study provides a reference for the performance evaluation of magnetostrictive composites before the preparation process.