Technical Diagnostics and Non-Destructive Testing, 2017, №01



2017 №01 (06)

DOI of Article 10.15407/tdnk2017.01.07

2017 №01 (08)

---

Technical Diagnostics and Non-Destructive Testing, №1, 2017 pp. 43-46

Realization of neural network algorithms for classification of technical state of composite materials by acoustic testing results

Galagan R. M., Momot A. S.

NTUU Igor Sykorsky Kiev Polytechnic Institute». 37, Peremohy Prosp.
E-mail: rgalagan@ukr.net

Possibility of application of APT-2 neural network architecture for development of algorithms for classification of technical condition of composite materials by the results of acoustic testing by free oscillations method, is considered. Interface and results of operation of a program that implements the APT-2 network are described, and a device for testing by free oscillations method was developed. 10 References, 2 Figures.
 
Keywords: neural networks, composite materials, defect classes, interface, diagnostics

References

1. Vorobey V. V., Markin V. B. Kontrol kachestva izgotovleniya i tekhnologiya remonta kompozitnykh konstruktsy. – Novosibirsk: Nauka, 2006. – 190 s. [in Russian].
2. Barynin V. A., Budadin O. N., Kulkov A. A. Sovremennye tekhnologii nerazrushayushchego kontrolya konstruktsy iz polimernykh kompozitsionnykh materialov. – M.: Spektr, 2013. – 242 s. [in Russian].
3. Klyuyev V. V., Yermolov I. N., Lange Yu. V. Nerazrushayushchy kontrol. Spravochnik v 7 t. – M: Mashinostroyeniye, 2004. – 864 s. [in Russian].
4. Barkhatov V. A. (2006) Recognizing Imperfections with an Artificial Neural Network of a Special Type. Russian Journal of Nondestructive Testing, #2, 92-100. https://doi.org/10.1134/S1061830906020045
5. Tekhnologii shtuchnikh neironnikh merezh: http://www. victoria.lviv.ua/html/neural_nets/zmist.htm.
6. Laurene V. Fausett Fundamentals of Neural Networks: Architectures, Algorithms, and Applications. Prentice-hall, 1994. – 461 p.
7. Barsky A. B. Neyronnye seti: raspoznavaniye, upravleniye, prinyatiye resheny. – M.: Finansy i statistika, 2004. – 176 s. [in Russian].
8. Kryuchin O. V., Zenkova N. A. (2011) Ispolzovaniye iskusstvennykh neyronnykh setey dlya resheniya zadach klassifikatsii na primere modelirovaniya meditsinskogo obyekta. Vestnik TGU, 3, 48-51.
9. Smolentsev N. K. Osnovy teorii veyvletov. Veyvlety v Matlab. – M.: DMK, 2005. – 304 s. [in Russian].
10. Carpenter G. A., Grossberg S. (1987) ART 2: self-organization of stable category recognition codes for analog input patterns. Applied Optics, 26. 4919-4930. https://doi.org/10.1364/AO.26.004919