Abstract
In this work we present a possible approach to predict the magnetic behavior of electric steels using a vector hysteron model implemented in a ``moving'' way, because the position of the elementary magnetic particles used to represent the magnetic material changes in function of the magnetization state. The moving functions are identified using a suitable set of experimental data and a numerical procedure based on a Metric-Topological-Evolutionary optimization. The accuracy and effectiveness of the method proposed are proved by comparison between computed and measured data related to circular magnetic induction loops at different amplitudes from the virgin state to the saturation region.