Increase of identification quality on the basis of genetic adjustment of models parameters for description non-stationary objects

Authors

A.R Akhatov, Doctor of Technical Sciences, Professor of the Department of Information Technology Samarkand State University, Tel.: +998902716418 (м.),Follow
N.O Isroilov, Student of the Department of Applied Mathematics and Informatics of Samarkand State University. Tel.: +998905208080 (м.),Follow

Abstract

Paper formulated the problem of increasing the quality of identification for casual time series (CTS), representing non-stationary objects in multicontour control systems on the basis of effective search of global and local extremums when parameters of target optimization function are rational. The new approach is offered to optimization of search by consideration statistical and dynamic characteristics of CTS as specimen of population generation, and by reflection of coordinate points set in space of evolution toward optimal values of target function. The modified genetic algorithm (GA) is developed on the basis of overlapping, generalization of opportunities and use of properties of genetic operators for evolutionary modeling. Procedures are entered for statistical selection of search points population, exception "unsuccessful" descendants, change of specimen generations on the basis of elimination and replacement unpromising specimen. Efficiency of the program complex to CTS identification is investigated on the basis of mechanisms to adjustment of parameters by recurrent dependences, traditional and modified GA.

First Page

168

Last Page

174

References

1. A.A.Barsegyan, M.S.Kupriyanov, V.V.Stepanenko, I.I.Holod, “Tehnologiya analiza danny'h: Data Mining, Visual Mining, Text Mining” [Data analysis technology: : Data Mining, Visual Mining, Text Mining], Sank Peterburg: BHV - Peterburg, 2007, 384 p. (in Russian)

2. J.Kennedy, R.C.Eberhart, “Particle swarm optimization” In Proceedings of IEEE International Conference on Neural Networks, pp. 1942-1948, 1995.

3. R.C.Eberhart, J Kennedy. “A new optimizer using particle swarm theory” Proceedings of the Sixth International Symposium on Micro Machine and Human Science MHS’95, pp. 39-43, 1995.

4. D.Rutkovskaya, M.Pilin'skiy, L.Rutkovskiy, “Neyronny'e seti, geneticheskie algoritmy' i nechetkie sistemy” [Neural networks, genetic algorithms, and fuzzy systems]'. Moskva: Goryachaya liniya, Telekom, 2004, 452 p. (in Russian)

5. K.V.Mahotilo, S.N.Petrashev, S.A.Sergeev “Geneticheskie algoritmy', iskusstvenny'e neyronny'e seti i problemy' virtual'noy real'nosti” [Genetic algorithms, artificial neural networks, and virtual reality problems], Har'kov: Osnova, 1997, 112 p. . (in Russian)

6. I.I.Jumanov, Z.T.Bekmurodov, “Identifikaciya sluchayny'h vremenny'h ryadov na osnove neyro-nechetkoy seti dlya povy'sheniya dostovernosti prognoza” [Identification of random time series or the basis of a neuro-odd network to increase the reliability of the forecast] HI Mejdunarodnaya Aziatskaya shkola «Problemy' optimizacii slojny'h sistem», Issy'k-Kul'skaya oblast', s. Bulan-Sogotu, pp. 258-264, 2015. . (in Russian)

7. O.I.Djumanov, “Sistema intellektual'nogo analiza i obrabotki danny'h na osnove geneticheskih algoritmov kontrolya dostovernosti izobrajeniy neprery'vny'h ob`ektov” [Data mining and processing system based on genetic algorithms for controlling the reliability of images of continuous objects], HI Mejdunarodnaya Aziatskaya shkola «Problemy' optimizacii slojny'h sistem», Issy'k-Kul'skaya oblast', s. Bulan-Sogotu, pp. 249-254, 2014. (in Russian).

8. O.I.Djumanov, “Metody' analiza sluchayny'h vremenny'h processov v sistemah upravleniya nestacionarny'mi processami” [Methods for analyzing stochastic processes in non-stationary process control systems], Respublikanskaya nauchnaya konferenciya «Perspektivy' razvitiya informacionny'h tehnologiy i telekommunikacionny'h sistem», 13-14 mart 2014 g., TUIT, chast' 1, Tashkent, 2014 , pp. 332-334. (in Russian)

9. O.I.Djumanov, “Optimizaciya obucheniya neyrosetevy'h sistem obrabotki danny'h na osnove dinamicheskih harakteristik” [Optimization of training of neural network data processing systems based on dynamic characteristics], «Problemy' informatiki i e`nergetiki», no. 3, pp. 63-70, 2011. (in Russian).

Recommended Citation

Akhatov, A.R and Isroilov, N.O (2018) "Increase of identification quality on the basis of genetic adjustment of models parameters for description non-stationary objects," Chemical Technology, Control and Management: Vol. 2018: Iss. 1, Article 28.
DOI: https://doi.org/10.34920/2018.1-2.168-174