Науковий вісник НГУ

Development and identification of the refined mathematical model of catalytic reforming

User Rating:  / 0

Details

Category: Electrical engineering

Last Updated on Friday, 09 May 2014 18:02

Published on Tuesday, 16 April 2013 11:28

Hits: 3891

Authors:

I.L. Levchuk, State Higher Educational Institution "Ukrainian State University of Chemical Technology", Lecturer of the Department of Computer Integrated Technology and Metrology, Dnepropetrovsk, Ukraine

Abstract:

Purpose. To develop an updated mathematical model of catalytic reforming, taking into account the non-stationarity of the described object and specific chemical transformations of the reacting mixture in separate reactors of the reactor block.

Methodology. The methods of simulation and mathematical modeling, decomposition and mathematical apparatus of neural networks made the basis of the research.

Findings. The new updated mathematical model of the catalytic reforming has been developed. Its verification with use of the experimental data obtained from the catalytic reforming process at Kremenchug Refinery showed that the error of the main calculated parameters decreased by 3–7%. Also the total time of identification reduced significantly (by 80% in some cases).

Originality. The mathematical model of the catalytic reforming which uses the principle of decomposition and examines the process as a set of three connected models of individual reactors with individual tuning parameters has been created. The approach allows us to use the features of the mechanism and kinetics of chemical transformations of the reaction mixture in each reactor, as well as to compensate at the identification phase the process non-stationarity due to the different rate of catalyst deactivation in the individual reactors of the reactor block. We have proposed the combined intellectual method of identification of the mathematical model based on the thermodynamic parameters which allow minimizing search time of the tuning coefficients by using the neural network to reduce the number of steps of iterative algorithm on the various stages of the identification.

Practical value. Catalytic reforming is the important stage of obtaining the high-octane component of motor fuels, as well as individual aromatic hydrocarbons (benzene, toluene, xylene) used in the petrochemical industry, and hydrogen. The efficiency of the catalytic reforming is largely determined by the quality of control. The development of the simple fast and sufficiently accurate mathematical models suitable for utilization in modern systems of the optimal control is an urgent problem.

References:

1. Слинько М.Г. Основы и принципы математического моделирования каталитических процессов / Слинько М.Г. – Новосибирск: Ин-т катализа им. Г.К. Борескова СО РАН, 2004. – 488 с.

Slinko, M.G. (2004), Osnovy i printsypy matematicheskogo modelirovaniya [Fundamentals and Principles of Mathematical Modeling of Catalytic Processes], Institute of Catalysisname of G.N. Boreskov CO RAS, Novosibirsk, Russia.

2. Цыпкин Я.З. Информационная теория идентификации / Цыпкин Я.З. – М.: Наука, 1995. – 336 с.

Tsypkin, Ya.Z. (1995), Informatsyonnaya teoriya identifikatsyi [Information Theory of Identification] Nauka, Moscow, Russia.

3. Головко В.А. Нейронные сети: обучение, организация и применение / В.А. Головко, А.И. Галушкин – М.: ИПРЖР, 2001. – 256 с.

Golovko, V.A. and Galushkin, A.I. (2001), Neyronnye seti: obucheniye, organizatsya i primeneniye [Neural Networks: Training, Organization and Use], IPRZhR, Moscow, Russia.

4. Ахметов С.А. Технология и оборудование процессов переработки нефти и газа: учеб. пособие / Под ред. С.А. Ахметова – Недра, 2006. – 868 с.

Akhmetov, S.A. (2006), Tekhnologiya i oborudovaniye protsessov pererabotki nefti i gaza [Technology and Equipment in Oil and Gas Processing], tutorial, Nedra.

2013_2_levchuk
2014-05-06  646.79 KB  1243

• < Prev
• Next >