Purpose – To devise a parametric study using a new application of the boundary element method (BEM) and to propose an efficient approach for speeding up the computation time of the BEM based on neural networks (NNs).Design/methodology/approach – A 3D finite elements formulation is combined with radial basis function NNs to speeding up the computation time.Findings – The paper shows how to estimate the role of thin slabs filled with unconventional media in order to increase the coupling values when placed between two metallic strips in a coupled microstrip line layout or to improve the shielding properties when used as absorber.Research limitations/implications – The numerical results here presented are not bianisotropic but can be easily extended to take into account bianisotropic media.Originality/value – The formulation is one of the only with the potential for investigating unconventional bianisotropic media like Chiral materials which are seen as one possible route to achieving double negative media.
[1]
D. Pozar.
Microwave Engineering
,
1990
.
[2]
Michel Nakhla,et al.
A neural network modeling approach to circuit optimization and statistical design
,
1995
.
[3]
Ganesh Kothapalli.
Artificial neural networks as aids in circuit design
,
1995
.
[4]
Alessandro Toscano,et al.
Analysis of microstrip antennas using neural networks
,
1997
.
[5]
M.C.E. Yagoub,et al.
A robust algorithm for automatic development of neural network models for microwave applications
,
2001,
2001 IEEE MTT-S International Microwave Sympsoium Digest (Cat. No.01CH37157).
[6]
S. Ramakrishna,et al.
Physics of negative refractive index materials
,
2005
.