Determination of position and orientation of conducting rod using a neural network
The problem of electromagnetic wave diffraction by a conducting rod has been solved using the integral equation technique. The diagram of backward scattering has been plotted. A wavelet packet decomposition of the obtained characteristic has been carried out. Based on the values of Shannon entropy for decomposition components we have constructed the feature vector. We have trained a neural network based on radial basis elements, which allows one to determine the position of a rod and its spatial orientation based on the feature vector. Numerical simulation and statistical analysis of the numerical results have been carried out.
SOLDATOVA, O.P.; GARSHIN, A.A. Convolutional neural network applied to handwritten digits recognition. Computer Optics, 2010, v.34, n.2, p.252, http://www.computeroptics.smr.ru/KO/PDF/KO34-2/13.pdf.
HAYKIN, S. Neural Networks: A Comprehensive Foundation. Upper Saddle River, New Jersey: Prentice Hall, 2009.
COLTON, D.L.; KRESS, R. Integral Equation Methods in Scattering Theory. New York: John Wiley & Sons, 1983, DOI: http://dx.doi.org/10.1002/zamm.19850650104.
PUN, CHI-MAN; LEE, MOON-CHUEN. Log-polar wavelet energy signatures for rotation and scale invariant texture classification. IEEE Trans. Pattern Analysis Machine Intelligence, May 2003, v.25, n.5, p.590-603, DOI: http://dx.doi.org/10.1109/TPAMI.2003.1195993.
GIBSON, WALTON C. The Method of Moments in Electromagnetics. New York: CRC Press, 2008, 272 p.
VOLAKIS, JOHN L.; SERTEL, KUBILAY. Integral Equation Methods for Electromagnetics. Ohio: SciTech, 2012, 408 p.
SADIKU, MATTHEW N.O. Numerical Techniques in Electromagnetics, 2nd ed. New York: CRC Press, 2001, 750 p.
MITTRA R. (ed.), Computer Techniques for Electromagnetics. NY: Pergamon Press, 1973.
DAUBECHIES, I. Ten Lectures on Wavelets. SIAM, 1992.
CHUI, C.K. An Introduction to Wavelets. Academic Press, 1992.
GOLDSTEIN, L.D.; ZERNOV, N.V. Electromagnetic Fields and Waves. Moscow: Sov. Radio, 1971 [in Russian].