Radiotekhnika
Publishing house Radiotekhnika

"Publishing house Radiotekhnika":
scientific and technical literature.
Books and journals of publishing houses: IPRZHR, RS-PRESS, SCIENCE-PRESS


Тел.: +7 (495) 625-9241

 

Selection and analysis of neural networks for problems of diagnosing electromechanical systems

Keywords:

O.V. Veselov – Dr.Sc.(Eng.), Professor, Department of Mechatronics and Electronic Systems of Vehicle, Vladimir State University named after A.&N. Stoletovs
V.M. Perepyolkin – Post-graduate Student, Department of Mechatronics and Electronic Systems of Vehicle, Vladimir State University named after A.&N. Stoletovs


The article discusses the possibility of using neural networks in diagnostics problems. The object of diagnosis is electromechanical systems. Based on the analysis results, three network architectures are selected. A probabilistic network, a counter-propagation net-work, and a perceptron are considered. For each of the networks, the number of neurons in the input layer, the number of layers, the training method, the activation function, the volume of the training sample, the number of training cycles were analyzed. The error and recognition speed were monitored. The ability of the network to recognize signals with noisy data is estimated. When using an analog-to-digital converter, the recognition results were evaluated when the bit width was changed. The width limits are established, at which we obtain satisfactory results. The reliability of the performed diagnostics was checked by increasing the input information flows. The proba-bilistic network is finally chosen.
Presents findings reflect the direction of further studies to improve the qualitative characteristics of the diagnosis procedure.

References:
  1. Uossermen F. Nejrokomp’yuternaya texnika: teoriya i praktika. M.: Mir. 1992. 184 s.
  2. NeuralNetworkToolbox– Matematika // MATLAB. Exponenta: E’lektronny’j resurs. URL = http://matlab.exponenta.ru/neuralnetwork/index.php (data obrashheniya: 10.01. 2016 - 03.09.2017).
  3. Rudakova T.A., Chervyakov N.I., Lubenczov V.F. Sravnenie i vy’bor nejrosetevy’x modelej na osnove metoda analiza ierarxij // Pribory’ i sistemy’ upravleniya, kontrol’, diagnostika. 2008. № 5. S. 1−5.
  4. Potyomkin V.G. Spravochnik po MATLAB. URL = http://matlab.exponenta.ru/ml/book2/index.php (data obrashheniya: 20.07.2016).
  5. Cherny’x I.V. Simulink: Instrument modelirovaniya dinamicheskix sistem. URL = http://matlab.exponenta.ru/simulink/book1 (data obrashheniya: 30.07.2016).
  6. German-Galkin S.G. Komp’yuternoe modelirovanie poluprovodnikovy’x sistem v MATLAB 6.0: Ucheb. posobie. SPb.: KORONA-print. 2001.
  7. Gul’tyaev A.K. Vizual’noe modelirovanie v srede MATLAB: Ucheb. kurs. SPb.: Piter. 2000.
June 24, 2020
May 29, 2020

© Издательство «РАДИОТЕХНИКА», 2004-2017            Тел.: (495) 625-9241                   Designed by [SWAP]Studio