Analysis of quadruple-ridged square waveguide by multilayer perceptron neural network model

Author

Tang, Yiming ; Zhao, Jianzhong ; Wu, Wen

Author_Institution

Nanjing Univ. of Sci. & Technol., Nanjing

fYear

2006

fDate

12-15 Dec. 2006

Firstpage

1912

Lastpage

1918

Abstract

Multilayer perceptron neural network (MLPNN) model is established to determine the characteristic parameters of quadruple-ridged square waveguides. The model is trained with backward propagation (BP) algorithm. Results obtained by MLPNN model are in good agreement with those available data. A design example is then presented for the quadruple-ridged square waveguide operating at 18-40 GHz. The proposed MLPNN model is very simple, fast and valid for designing quadruple-ridged square waveguides and other microwave devices.

Keywords

electrical engineering computing; microwave propagation; millimetre wave propagation; multilayer perceptrons; waveguides; backward propagation algorithm; microwave devices; multilayer perceptron neural network model; quadruple-ridged square waveguide; waveguide devices; Algorithm design and analysis; Design automation; Equations; Impedance; Magnetic analysis; Microwave propagation; Microwave technology; Multi-layer neural network; Multilayer perceptrons; Neural networks; backward propagation (BP); multilayer perceptron(MLP); neural network model(NNM); quadruple-ridged; square waveguide;

fLanguage

English

Publisher

ieee

Conference_Titel

Microwave Conference, 2006. APMC 2006. Asia-Pacific

Conference_Location

Yokohama

Print_ISBN

978-4-902339-08-6

Electronic_ISBN

978-4-902339-11-6

Type

conf

DOI

10.1109/APMC.2006.4429782

Filename

4429782