Preview

Advanced Engineering Research (Rostov-on-Don)

Advanced search

ARTIFICIAL NEURAL NETWORK FAULT MODEL STUDY

Abstract

Fault models of the artificial neural network hardware implementations are considered. The development of the neuroelement fault classes through the introduction of the additional neuron fault model class causing certain level signals appearance on their output in the interval (0,1) is offered. The experimental research of the introduced fault class impact on the artificial neural network operability is done.

About the Authors

Vladimir A. Fatkhi
Don State Technical University.
Russian Federation


Daniil V. Marshakov
Don State Technical University.
Russian Federation


Vasily V. Galushka
Don State Technical University.
Russian Federation


References

1. Himavathi S. Feedforward Neural Network Implementation in FPGA Using Layer Multiplexing for Effective Resource Utilization / S. Himavathi, D. Anitha, A. Muthuramalingam // IEEE Transactions on Neural Networks. – 2007. – Vol. 18. – №3. – pp. 880-888.

2. Gribachyov V.P. E`lementnaya baza apparatny`x realizacij nejronny`x setej / V.P. Gribachyov // Komponenty` i texnologii. – 2006. – №8. – S. 12-15. – In Russian.

3. Xajkin S. Nejronny`e seti: polny`j kurs, 2-e izdanie: per. s angl. / S. Xajkin. – M.: Izdatel`skij dom «Vil`yams», 2006. – 1104 s. – In Russian.

4. Osovskij S. Nejronny`e seti dlya obrabotki informacii / S. Osovskij. – M.: Finansy` i statistika, 2002. – 344 s. – In Russian.

5. Bolt G. Fault models for artificial neural networks / G. Bolt // IEEE International Joint Conference on Neural Networks, Singapore, 1991. – pp. 1373-1378.

6. Phatak D.S. Complete and Partial Fault Tolerance of Feedforward Neural Nets / D.S. Phatak // IEEE Transactions on Neural Networks, 1995. – Vol. 6/2. – pp. 446-456.

7. Piuri V. Fault tolerance in neural networks: theoretical analysis and simulation results / V. Piuri, M. Sami, R. Stefanelli // Advanced Computer Technology, Reliable Systems and Applications, 5th Annual European Computer Conference, Bologna, Italy, 1991. – pp. 429-436.

8. Marshakov D.V. Issledovanie vliyaniya defektov na rabotosposobnost` mnogoslojnoj is-kusstvennoj nejronnoj seti pryamogo rasprostraneniya / D.V. Marshakov, V.A. Fatxi // Vestnik DGTU. – 2011. – T. 11. – №2 (53). – S. 169-173. – In Russian.

9. Dias F.M. Fault Tolerance of Artificial Neural Networks: an Open Discussion for a Global Model / F.M. Dias, A. Antunes // International Journal of Circuits, Systems and Signal Processing, Naun, July, 2008. – pp. 9-16.

10. Parxomenko P.P. Osnovy` texnicheskoj diagnostiki. (Optimizaciya algoritmov diagnostiro-vaniya, apparaturny`e sredstva) / P.P. Parxomenko, E.S. Sogomonyan. – M.: E`nergiya, 1981. – 320 s. – In Russian.

11. Marshakov D.V. Ranzhirovanie defektov v iskusstvenny`x nejronny`x setyax / D.V. Mar-shakov, V.A. Fatxi // Matematicheskie metody` v texnike i texnologiyax – MMTT23: sb. tr. XXIII mezh-dunar. nauch. konf.: v 12 t. – Saratov, 2010. – T. 5. – Sekciya 5. – S.210-211. – In Russian.

12. Singh A.P. Empirical study of FFANN tolerance to weight stuck-at max/min fault / A.P. Singh, C.S. Rai, P. Chandra // Internation Journal of Artificial Intelligence & Applications (IJAIA), 2010. – Vol. 1. – №2. – pp. 13-21.

13. Omondi A. FPGA Implementations of Neural Networks / A. Omondi, J. Rajapakse, M. Bajger. – Springer, Berlin, Germany, 2006. – 360 p.

14. Galushkin A.I. Nejronny`e seti: osnovy` teorii / A.I. Galushkin. – M.: Goryachaya liniya-Telekom, 2010. – 480 s. – In Russian.


Review

For citations:


Fatkhi V.A., Marshakov D.V., Galushka V.V. ARTIFICIAL NEURAL NETWORK FAULT MODEL STUDY. Vestnik of Don State Technical University. 2012;12(3):65-71. (In Russ.)

Views: 402


Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.


ISSN 2687-1653 (Online)