Artificial neural networks for interpolation and identification of underwater object features

• Autorzy: Balicki J. , Gloza I.
• Języki publikacji: EN

Abstrakty

EN

Artificial neural networks can be applied for interpolation of function with multiple variables. Because of concurrent processing of data by neurons, that approach can be seen as hopeful alternative for numerical algorithms. From these reasons, the analysis of capabilities for some models of neural networks has been carried out in the purpose for identification of the underwater object properties. Features of the underwater objects can be recognized by characteristics of a amplitude according to the frequency of measured signals. The feed-forward multi-layer networks with different transfer functions have been applied. Those network models have been trained by some versions of back-propagation algorithm as well as the Levenberg-Marquardt gradient optimization technique. Finally, for determination of the amplitude for the frequency of signal by the two-layer network with the hidden layer of the radial neurons has been proposed.

Słowa kluczowe

EN

artificial neural networks; underwater object; hydroacoustic

• Wydawca: Polskie Towarzystwo Akustyczne. Oddział Gdański
• Czasopismo: Hydroacoustics
• Rocznik: 2008
• Tom: Vol. 11
• Strony: 1--10
• Opis fizyczny: Bibliogr. 12 poz., rys.

Twórcy

autor

Balicki J.

autor

Gloza I.

• Naval University of Gdynia, ul. Śmidowicza 69, 81-103 Gdynia, Poland,

Bibliografia

• 1. J. Balicki, Artificial neural networks for shape modeling of sea bottom, Hydroacoustics, Vol. 10, 2007, pp. 1-8.
• 2. S. Chen, C.F. Cowman, P.M. Grant, Orthogonal least squares learning algorithm for radial basis function learning, IEEE Trans. Neural Networks, Vol. 2, 1991, pp. 302-309.
• 3. E.K.P. Chong, S.H. Zak, An introduction to optimization, John Wiley&Sons, Inc., New York, 1996.
• 4. S. Coombes, S. H. Doole, Neuronal Population Dynamics with Post Inhibitory Rebound: A Reduction to Piecewise Linear Discontinuous Circle Maps, Dynamics and Stability of Systems, Vol. 11, 1996, pp. 193-217.
• 5. H. Demuth, M. Beale, Matlab Neural Network Toolbox. User’s Guide, The MathWorks, Inc., 2003.
• 6. M. Frize, C. Ennett, M. Stevenson, H. Trigg, Clinical decision-support systems for intensive care units using artificial neural networks, Medical Engineering and Physics, 2001, Vol. 23, pp. 217-225
• 7. I. Gloza, Underwater Diagnostic Method of Propulsive Machinery, Acta Acustica united with Acustica, Vol. 92, 2005, pp. 156-158.
• 8. T. Kujanpää, J. Roos, Efficient Initialization of Artificial Neural Network Weights for Electrical Component Models, Book of Abstracts of SCEE 2006, Sinaia, Romania, 2006, pp. 47-48.
• 9. M. Mokhari, Classification by the Random Neural Network, Tenth International Conference on Mathematical and Computer Modelling and Scientific Computing, Boston, USA, 1995.
• 10. S. Priddy; P. E. Keller, Artificial Neural Networks: An Introduction, SPIE Press, New York 2005.
• 11. A.J.C. Sharkey, G.O. Chandroth, N.E. Sharkey, Acoustic Emisssion, Cylinder Pressure and Vibration: A Multisensor Approach to Robust Fault Diagnosis, In Proceedings of IJCNN2000, Como, Italy, 2000.
• 12. L. M. Silva, L. A. Alexandre, J. Marques de Sá., New developments of the z-edm algorithm, In 6th International Conference on Intelligent Systems Design and Applications (ISDA'06), volume 1, pp. 1067-1072, Jinan, China, October 2006.