|
Digital
Library of the European Council for Modelling and Simulation |
Title: |
A Robust And Adaptive Approach To Control Of A Continuous Stirred Tank
Reactor With Jacket Cooling |
Authors: |
Roman Prokop, Radek Matusu, Jiri Vojtesek |
Published in: |
(2021). ECMS 2021,
35th Proceedings DOI: http://doi.org/10.7148/2021 ISSN:
2522-2422 (ONLINE) ISSN:
2522-2414 (PRINT) ISSN:
2522-2430 (CD-ROM) ISBN: 978-3-937436-72-2 Communications of the ECMS , Volume 35, Issue 1, June 2021, United Kingdom |
Citation
format: |
Roman Prokop,
Radek Matusu, Jiri Vojtesek (2021). A Robust
And Adaptive Approach To Control Of A Continuous Stirred Tank Reactor With
Jacket Cooling, ECMS 2021 Proceedings
Edited By: Khalid Al-Begain, Mauro Iacono, Lelio Campanile, Andrzej Bargiela European
Council for Modeling and Simulation. doi:
10.7148/2021-0185 |
DOI: |
https://doi.org/10.7148/2021-0185 |
Abstract: |
Continuous
Stirred Tank Reactors (CSTR) are one of the main
technological plants used in chemical and biochemical industry. These systems
are quite complex with many nonlinearities and the conventional linear
control with fixed parameters can be questionable or sometimes unacceptable.
The solution should be found in so-called “non-traditional” control
approaches like adaptive, robust, fuzzy, or artificial intelligent methods.
One way is the utilization of self-tuning adaptive schemes, but computations
may be quite difficult, clumsy and time-consuming. This paper brings an
alternative principle called a robust approach and the comparison of the
robust and adaptive control responses. Robust control considers a CSTR model
as a linear system with parametric uncertainty, which covers a family of all
feasible plants. Then several controllers with fix parameters are designed so
that for all possible plants, the acceptable control behavior is obtained.
The two-degree-of-freedom (2DOF) structure for the control law was chosen.
Both robust and adaptive control is applied to an original nonlinear model of
a CSTR. All calculations and simulations of mathematical models and control
responses were performed in the Matlab and Simulink environment. |
Full
text: |