Applications of BPNN in Analyzing LHM Elemental Basic Structure

Article Preview

Abstract:

Based on the back propagation multi-layer forward feed neural network, the neural network model is built for the dielectric sensitive structural parameters between the equivalent permittivity and the equivalent permeability, which is used to analyze the basic left-hand materials(LHMS) structural. The experimental results show that the analysis time is 145.535648 seconds and the training mean error is 0.000113426 while adopting the scaled conjugate gradient method. The results are coincident with these ones by the full wave method, satisfying the engineering demand, reducing the faults caused by thickness resonance in the traditional numerical analysis method, realizing the coexistence between the high analysis precision and the high efficiency of the left-hand materials.

You might also be interested in these eBooks

Info:

Periodical:

Pages:

466-470

Citation:

Online since:

October 2009

Export:

Price:

Permissions CCC:

Permissions PLS:

Сopyright:

© 2009 Trans Tech Publications Ltd. All Rights Reserved

Share:

Citation:

[1] L.X. Ran, H. Sh. Chen and J.T. Hung: Submitted to Journal of MicroWaves (2004).

Google Scholar

[2] P. Pan, Q. Wu and F.Y. Meng: Submitted to Journal of Telemetry, Tracking and Command (2007).

Google Scholar

[3] D.R. Smith, W.J. Padilla and D.C. Vier: Submitted to Physical Review (2000).

Google Scholar

[4] H. SH. Chen: Study of Equivalent Circuit Theory and Experimentation for Left-handed Material, (Zhejiang University Publishing, Zhejiang 2005).

Google Scholar

[5] Q. Wu and M.F. Wu: Submitted to Chinese Journal of Radio Science (2006).

Google Scholar

[6] L. Rong, A.G. Wang, Y. Lv: Submitted to Journal of Micro Waves (2007).

Google Scholar

[7] H.K. Wei: Theory and Methods of !eural !etwork Structural Design (National Defence Industry Press, 2005).

Google Scholar

[8] Y. Liu and L.Y. Zhang: Submitted to Electronic Measurement Technology (2007).

Google Scholar

[9] Moller and Martin. Fodslette: Submitted to Neural Networks (2007).

Google Scholar