Method and algorithms of nonlinear dynamic processes identification

Authors:

V.I. Korniіenko, Dr. Sci. (Tech.), State Higher Educational Institution “National Mining University”, Professor of Security Information and Telecommunications Department, Dnipropetrovsk, Ukraine.

S.M. Matsiuk, State Higher Educational Institution “National Mining University”, Doctoral Student of the Department of Computer Systems Software, Dnipropetrovsk, Ukraine.

I.M. Udovyk, Cand. Sci. (Tech.), State Higher Educational Institution “National Mining University”, Docent of the Department of Computer Systems Software, Dnipropetrovsk, Ukraine.

О.M. Alekseіev, Cand. Sci. (Tech.), State Higher Educational Institution “National Mining University”, Senior Lecturer of the Department of System Analysis and Control, Dnipropetrovsk, Ukraine.

Abstract:

Purpose. Increasing accuracy of dynamic models of the complex nonlinear processes for solution of the tasks for these processes control.

Methodology. Structural-parametric identification of nonlinear dynamic processes including the identification of the mo-del structure based on selection by non-shift criterion, as well as self-reactance identification of the optimum structure model by the regularity criterion through the whole sampling of experimental data.

Findings. The algorithms of global and local optimization of nonlinear dynamic process models realizing the procedure of structural-parametric identification by their structural and self-reactance optimization were developed, which allows getting the models of extra accuracy.

Originality. The method of identification of nonlinear dynamic processes consisting of procedures for estimation of the state and characteristics of the process, as well as their structural-parametric identification, was offered. It allows, unlike the known of methods, to fulfil the identification of these processes in the batch mode by structural-parametric and in real-time mode by self-reactance optimization of their models.

Practical value. The results of the research can be used while developing algorithms for controlling complex nonlinear processes based on their complex estimation and identification.

References:

1. Kоrnienko, V.I., Gulina, I.G. and Budkova, L.V., 2013. Complex estimation, identification and prediction of difficult nonlinear processes. Naukovyi Visnyk Natsionalnoho Hirnychoho Universytetu, 6, pp. 124−131.

Корнієнко В.І. Комплексна оцінка, ідентифікація та прогнозування cкладних нелінійних процесів / В.І. Корнієнко, І.Г. Гуліна, Л.В. Будкова // Науковий вісник НГУ. – 2013. – № 6. − С. 124−131

2. Kuznetsov, G.V., Kornienko, V.I. and Gerasina, O.V., 2009. Composition strucrural-parametric identification of nonlinear dynamic controlled objects. Naukovi Visti NTUU “KPI”, 5, pp. 69−75.

Кузнецов Г.В. Композиційна структурно-параметрична ідентифікація нелінійних динамічних об’єктів керування / Г.В. Кузнецов, В.І. Корнієнко, О.В. Герасіна // Наукові вісті НТУУ КПІ. – 2009. – № 5. – С. 69−75.

3. Kuznetsov, S.P., 2002. Dinamicheskiy khaos [Dynamic chaos]. Moscow: Fizmatlit.

Кузнецов С.П. Динамический хаос / Кузнецов С.П. – М.: Физматлит, 2002. – 296 с.

4. Voronovskiy, G.K. and Makhotilo, K.V., 1997. Geneticheskie algoritmy, iskusstvennye neyronnye seti i problemy virtualnoy realnosti [Genetic algorithms, artificial neuron networks and problems of virtual reality]. Kharkov: Osnova.

Вороновский Г.К. Генетические алгоритмы, искусственные нейронные сети и проблемы виртуальной реальности / Вороновский Г.К., Махотило К.В. – Харьков: Основа, 1997. – 112 с.

5. Kruglov, V.V., Dli, M.I. and Golunov, R.Yu., 2001. Nechetkaya logika i iskysstvennye neyronnye seti [Fuzzy logic and artificial neuron networks]. Moscow: Fizmatlit.

Круглов В.В. Нечеткая логика и искусственные нейронные сети /. Круглов В.В, Дли М.И., Голунов Р.Ю. – М.: Физматлит, 2001. – 224 с.

2016_01_korniienko
2016-04-02  487.08 KB  860