Title :
Estimation of thickness of subsurface air layer by neuron network technology application to reflected microwave signal
Author :
Drobakhin, O.O. ; Doronin, A.V.
Author_Institution :
Dept. of Phys., Dnepropetrovsk Nat. Univ., Dnipropetrovsk
fDate :
June 29 2008-July 2 2008
Abstract :
Neuron network technology application to subsurface air layer thickness estimation is considered. The envelope of time-domain signal was used as input data. The three-layered neuron network with backpropagation and sigmoid (S-shaped) activation function of neurons was chosen. The comparison with results of correlation method is presented.
Keywords :
backpropagation; correlation methods; neural nets; nondestructive testing; correlation method; neuron backpropagation; neuron network; nondestructive testing; reflected microwave signal; sigmoid activation function; subsurface air layer; thickness estimation; time-domain signal; Application software; Artificial neural networks; Backpropagation; Correlation; Electromagnetic reflection; Frequency; Geophysical measurements; Microwave technology; Neurons; Time domain analysis;
Conference_Titel :
Mathematical Methods in Electromagnetic Theory, 2008. MMET 2008. 12th International Conference on
Conference_Location :
Odesa
Print_ISBN :
978-1-4244-2284-5
DOI :
10.1109/MMET.2008.4580920
