Neural network based scalar hysteresis model

Issue title: Proceedings of the tenth International Symposium on Applied Electromagnetics and Mechanics ISEM-Tokyo

Article type: Research Article

Authors: Kuczmann, Miklós^{a; *} | Iványi, Amália^a

Affiliations: [a] Department of Electromagnetic Theory, Budapest University of Technology and Economics, Egry J. u. 18, Budapest, H-1521, Hungary

Correspondence: [*] Corresponding author: Miklós Kuczmann, Tel.: +36 1 463 2915; Fax: +36 1 463 3189; E-mail: kuczman@evtsz1.evt. bme.hu

Abstract: A neural network (NN) model of scalar hysteresis phenomena has been developed for modeling the behavior of magnetic materials. The function approximation ability of NNs has been applied. The virgin curve and a set of the first order reversal branches can be stored preliminary in a system of three NNs. Different properties of magnetic materials can be simulated by a knowledge-based algorithm. Value of differential susceptibility, dM/dH can be expressed in analytical form. Finally hysteresis characteristics predicted by the introduced model are compared with the results of the Preisach simulation technique.

DOI: 10.3233/JAE-2002-508

Journal: International Journal of Applied Electromagnetics and Mechanics, vol. 15, no. 1-4, pp. 225-230, 2002