Abstract : The purpose of this paper is to describe an improved microwave method for predicting the material’s thickness and the saturation magnetization and the damping factor through the neural networks. These characteristics provide the permeability tensor components using the combination between theoretical models and neural network. Neural networks learn the relationship between the scattering parameters and the outputs. The networks’ performances result from both simulation and measurement thin ferrite samples.
https://hal-ujm.archives-ouvertes.fr/ujm-01390559
Contributeur : Stéphane Robert
<>
Soumis le : mercredi 2 novembre 2016 - 10:55:11
Dernière modification le : jeudi 11 janvier 2018 - 06:20:36
