The aim of this work is to present a new tool for the analysis of magnetic field problems considering 2-D magnetic hysteresis. In particular, this tool makes use of the Finite Element Method to solve the magnetic field problem in real device, and fruitfully exploits a neural network (NN) for the modeling of 2-D magnetic hysteresis of materials. The NS has as input the magnetic inductions components B at the k-th simulation step and returns as output the corresponding values of the magnetic field H corresponding to the input pattern. It is trained by vector measurements performed on the magnetic material to be modeled. This input/output scheme is directly implemented in a FEM code employing the magnetic potential vector A formulation. Validations through measurements on a real device have been performed.
A Neural-FEM tool for the 2-D magnetic hysteresis modeling / Cardelli E.; Faba A.; Laudani A.; Lozito G.M.; Riganti Fulginei F.; Salvini A.. - In: PHYSICA. B, CONDENSED MATTER. - ISSN 0921-4526. - ELETTRONICO. - 486:(2016), pp. 111-115. [10.1016/j.physb.2015.12.006]
A Neural-FEM tool for the 2-D magnetic hysteresis modeling
Lozito G. M.;
2016
Abstract
The aim of this work is to present a new tool for the analysis of magnetic field problems considering 2-D magnetic hysteresis. In particular, this tool makes use of the Finite Element Method to solve the magnetic field problem in real device, and fruitfully exploits a neural network (NN) for the modeling of 2-D magnetic hysteresis of materials. The NS has as input the magnetic inductions components B at the k-th simulation step and returns as output the corresponding values of the magnetic field H corresponding to the input pattern. It is trained by vector measurements performed on the magnetic material to be modeled. This input/output scheme is directly implemented in a FEM code employing the magnetic potential vector A formulation. Validations through measurements on a real device have been performed.I documenti in FLORE sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.