Defense of the dissertation of Boranbayeva Narkez for the degree of Doctor of Philosophy (PhD) in the specialty «8D07102 - Automation and control»

L.N. Gumilyov Eurasian National University, a dissertation defense for the degree of Doctor of Philosophy (PhD) by Boranbayeva Narkez on the topic «Control system for operating modes of the catalytic cracking unit in conditions of fuzzy initial information» to the educational program «8D07102 – Automation and control».

The dissertation was carried out at the «System analysis and management education department» of L.N. Gumilyov Eurasian National University.

The language of defense is kazakh

Official reviewers:

Temporary members of the Dissertation Committee:

Scientific advisors:

Orazbayev Batyrbai Bidaybekovich - Doctor of Technical Sciences, Professor of the Department of System Analysis and Control, L.N. Gumilyov Eurasian National University (Astana, Republic of Kazakhstan).

Wójcik Waldemar – Doctor of Technical Sciences, Professor of the Department of Electronics and Information Technology, Faculty of Electrical Engineering and Computer Science, Lublin University of Technology (Lublin, Poland).

The defense will take place on June 05, 2026, at 02:00 PM in the Dissertation Council for the training direction «8D071 – Engineering and engineering trades» in the specialty «8D07102 – Automation and control» of L.N. Gumilyov Eurasian National University. The defense meeting is planned to be held online.

Link: https://teams.microsoft.com/meet/43084063069554?p=sYYIBSTyZZz7EHsgT2

Address: Астана қ., Пушкин көшесі, 11, №2 ғимарат, Мәжіліс залы (№222 ауд.).

Abstract (English): ABSTRACT of the Dissertational work of the PhD Thesis on the educational programm 8D07102 - «Automation and control» Narkez Boranbayeva «Control System for Operating Modes of a Catalytic Cracking Unit under Conditions of Uncertain Initial Information». Relevance of the research topic. Catalytic cracking is one of the key processes in oil refineries, providing the production of high-octane components of automotive gasoline and other valuable products. This process plays a significant role in deep oil refining and is strategically important for the formation of the national fuel balance. At the Shymkent Oil Refinery, a Resid Fluid Catalytic Cracking (RFCC, unit 1000) facility is operated to convert heavy oil fractions into light hydrocarbons. The process is carried out under fluidized bed conditions, ensuring high intensity of mass and heat transfer and contributing to increased yields of target products. The main objective of the unit is to increase the depth of oil refining and the share of light petroleum products. In the reactor, heavy hydrocarbon molecules undergo cracking reactions, forming gasoline and gas fractions. In the regenerator, coke is burned off from the catalyst surface, enabling its reuse in the technological cycle. The efficiency of the unit is largely determined by the stability of the reactor–regenerator block, which is the core element of the technological system. The catalytic cracking process belongs to the class of complex, multi-parameter, and poorly formalized chemical-technological processes. It is characterized by a high degree of uncertainty in initial information, caused by variability in feedstock quality and operating conditions. Particular importance is associated with the problem of predicting gasoline quality, as it depends on numerous factors and is subject to stochastic disturbances. This significantly complicates the development of adequate mathematical models and control systems based on conventional approaches. One of the most promising directions in this field is the application of fuzzy logic and linguistic modeling methods, which enable the consideration of uncertainty in initial data, the integration of expert knowledge, and decision-making under conditions of incomplete information. The formulation of membership functions for linguistic variables based on the experience and intuition of domain experts creates the necessary prerequisites for improving the accuracy and reliability of process control. Under these conditions, the development of an intelligent control system for the operating modes of the reactor–regenerator block of a catalytic cracking unit represents a relevant scientific and practical problem. In industrial practice, a number of parameters and performance indicators of RFCC units are characterized by uncertainty. For such complex technological systems, determining optimal operating modes requires the consideration of fuzzy information in the form of expert knowledge, experience, and intuition. The development of intelligent systems capable of effectively utilizing such information is currently one of the most important scientific and practical challenges. Thus, the selected research topic, aimed at developing methods and algorithms for intelligent control of catalytic cracking under conditions of uncertainty, is highly relevant and possesses significant theoretical and practical importance. In the context of Industry 4.0 and the digital transformation of oil refining, the development of intelligent and adaptive control systems becomes increasingly important. Such systems enable real-time decision-making, improve operational efficiency, and enhance the overall reliability of technological processes. The purpose of the dissertation research to to develop an intelligent control system for the operating modes of a catalytic cracking unit based on fuzzy logic, mathematical modeling, and machine learning algorithms. To achieve this objective, the following tasks are addressed: - to analyze the current state of mathematical modeling and control methods for catalytic cracking units, identify existing approaches, and determine their limitations; - to investigate the technological features of the reactor–regenerator block of the catalytic cracking unit and determine the key process parameters affecting product yield and quality; - to develop mathematical models of the reactor–regenerator block that ensure an adequate description of process dynamics under conditions of multi-parameter complexity and uncertainty of initial data; - to develop an intelligent control system for catalytic cracking unit operating modes based on the integration of regression models, fuzzy logic, expert knowledge, and machine learning methods; - to perform experimental validation of the developed control system using the catalytic cracking unit of the Shymkent Oil Refinery as a case study and to evaluate its efficiency in comparison with conventional control methods. Object and Subject of Research. The object of the study is the technological process of catalytic cracking of heavy oil residues in the reactor-regenerator block of the Shymkent refinery. The subject of the study is mathematical models and methods for intelligent control of operating modes based on fuzzy set theory, expert knowledge, fuzzy modeling, and machine learning. The solution of these tasks ensures the development of a comprehensive intelligent control framework for catalytic cracking units operating under uncertainty. Research Methods. In the course of the dissertation research, a comprehensive set of modern research methods was employed, including the analysis of scientific literature and industrial data related to the problems of mathematical modeling and control of catalytic cracking processes. To study the dynamics and interrelationships of the parameters of the reactor–regenerator block of the catalytic cracking unit, methods of system analysis and mathematical modeling were applied, ensuring an adequate representation of the process behavior under conditions of multi-parameter complexity and uncertainty. In addition, data preprocessing techniques and validation methods were applied to ensure the reliability and robustness of the developed models. Scientific Novelty of the Dissertation Research. The scientific novelty of the dissertation research is defined by the following contributions: 1. A novel hybrid methodology for modeling the operating modes of the reactor–regenerator block of a residue fluid catalytic cracking (RFCC) unit has been developed and theoretically substantiated for the first time. The proposed methodology integrates classical regression models, fuzzy logic, and machine learning algorithms, enabling a comprehensive representation of the nonlinear, stochastic, and multi-parameter nature of the catalytic cracking process. It is theoretically justified and experimentally confirmed that the hybrid approach significantly improves the accuracy and robustness of the models. 2. A new structure of a fuzzy model based on the linguistic representation of the technological process has been proposed. While the model is built upon the classical Mamdani framework, its distinctive feature is the dynamic adaptation of the fuzzy rule base using machine learning techniques, which ensures continuous adjustment to changing operating conditions and automatic updating of model parameters in real time. 3. Based on industrial data from the Shymkent Oil Refinery and expert knowledge, an adaptive fuzzy model has been developed, incorporating the capability of automatic correction of expert rules using machine learning methods. The model captures the relationships between key parameters of the reactor–regenerator block and gasoline yield, thereby enhancing the accuracy of process description and predictive performance. 4. A novel structural and algorithmic framework for intelligent control of catalytic cracking unit operating modes has been developed. For the first time, an integrated application of fuzzy logic and machine learning modules within a unified control architecture is proposed, leading to improved decision-making accuracy and enhanced process stability. The effectiveness of the proposed intelligent control system has been experimentally validated using industrial data from the Shymkent Oil Refinery. 5. As a result of the research, intellectual property rights have been formally registered, including a methodology for intelligent control of the catalytic cracking process of heavy oil residues (National Institute of Intellectual Property of the Republic of Kazakhstan, No. 66362, 2026), a software algorithm for intelligent control of catalytic cracking unit operating modes under conditions of uncertain initial information (National Institute of Intellectual Property of the Republic of Kazakhstan, No. 66507, 2026). The proposed solutions contribute to the advancement of intelligent process control methods and expand the applicability of hybrid modeling approaches in complex industrial systems. Main Scientific Contributions Submitted for Defense. The following scientific contributions and results are submitted for defense in the dissertation: 1. A novel hybrid modeling methodology for the reactor-regenerator block of a catalytic cracking unit is proposed and theoretically substantiated for the first time. The methodology accounts for the nonlinear and multi-objective nature of the process. It is demonstrated that the proposed approach significantly improves modeling accuracy and enhances the efficiency of process control. 2. Based on industrial data from the Shymkent Oil Refinery, a system of relationships describing the interdependencies among key process parameters affecting gasoline yield and quality has been obtained. These relationships are formalized in the form of a hybrid fuzzy-polynomial model, providing an adequate description and reliable prediction of the technological process. 3. A new structural scheme and an algorithm for intelligent control of catalytic cracking unit operating modes are proposed. It is shown that the developed algorithm improves process stability, prediction accuracy, and the effectiveness of control decisions. Based on the proposed models, an intelligent control system for the catalytic cracking unit has been developed. 4. The adequacy of the developed models has been verified, and their experimental validation has been carried out using real industrial data from the Shymkent Oil Refinery. The effectiveness of the proposed intelligent control system is demonstrated, along with its advantages over conventional control methods. The experimental results show an increase in gasoline yield by 3.2%, an improvement in gasoline density by 0.4%, a reduction in process instability indicators. The obtained results confirm the effectiveness of the proposed approach and its applicability to real industrial systems. Practical Significance of the Research Results. The practical significance of the research results is as follows: 1. The developed control system enables the determination of optimal values of technological process parameters, ensuring increased gasoline yield and improved product quality. The proposed system can be implemented at oil refineries to enhance the efficiency of gasoline production. 2. The developed models provide more accurate prediction of process behavior under conditions of uncertainty and allow the formation of optimal operating modes of the unit. The proposed models and algorithms can be applied in the modernization of automated control systems for catalytic cracking units. 3. The research results can be utilized in the modernization of automated process control systems at oil refineries, as well as in the training of operators and engineering personnel, contributing to improved efficiency of equipment operation. 4. The main results of the dissertation research have been implemented in the educational process at L.N. Gumilyov Eurasian National University within the educational program 6B07102 – “Automation and Control”, and are used in teaching and research activities. The implementation of the proposed system can lead to economic benefits through increased efficiency, reduced operational risks, and improved product quality. Compliance with Scientific Development Priorities and State Programs. The dissertation research corresponds to one of the priority areas of scientific development of the Republic of Kazakhstan, namely «Advanced Manufacturing, Digital and Space Technologies». The research presented in this dissertation was carried out within the framework of grant funding for young scientists, in accordance with Contract No. 52ЖҒ-25-27 dated February 27, 2025, under the project AP25795091 entitled «Development of an Intelligent Control System for the Catalytic Cracking Process». Author’s Contribution. The author’s contribution to the obtained scientific results is as follows: 1. A comprehensive analysis of the current state of mathematical modeling and control of catalytic cracking processes was carried out, and both domestic and international studies in this field were systematized and generalized. 2. The technological features of the reactor–regenerator block of the catalytic cracking unit at the Shymkent Oil Refinery were investigated, and the key parameters affecting product yield and quality were identified. 3. Mathematical models of the reactor–regenerator block were developed, taking into account the multi-parameter nature of the process and the uncertainty of initial data. 4. A method for the synthesis of linguistic models of technological systems based on expert knowledge and fuzzy set theory was proposed and implemented. 5. Fuzzy models of the reactor–regenerator block were developed and implemented using MATLAB (Fuzzy Logic Toolbox) and Python programming environments. 6. The structure and software implementation of an intelligent control system for the operating modes of the catalytic cracking unit were developed. 7. Verification and experimental validation of the proposed models and control system were carried out using industrial data from the Shymkent Oil Refinery. 8. Scientific statements reflecting the novelty of the work were formulated, and directions for practical application of the obtained results were identified. All stages of the research, including model development, implementation, and validation, were carried out by the author. Approbation of Research Results and Publications. The main results of the dissertation research have been published in 6 scientific papers, including 3 articles in journals recommended by the Committee for Quality Assurance in Science and Higher Education of the Ministry of Science and Higher Education of the Republic of Kazakhstan: 1. Development of a mathematical model of the reactor–regenerator block of a catalytic cracking unit // Bulletin of KazATC – 2024. – Vol. 1, No. 130. – pp. 153–162. 2. Forecasting and optimization of catalytic cracking unit operation under condition of fuzzy information // Scientific Journal of Astana IT University – 2024. – Vol. 19. – pp. 46–59. 3. Determination of product yield of a catalytic cracking unit using the Python programming environment // Bulletin of KazUTB – 2024. – Vol. 4, No. 25. – pp. 84–96. The results of the dissertation research have also been presented at international scientific conferences: 1. An approach to modeling a catalytic cracking unit based on intelligent methods // “Creativity of Youth for Innovative Development of Kazakhstan”: Proceedings of the XX Scientific and Technical Conference, 2024. – pp. 54–58. The main scientific results of the dissertation have been published in journals indexed in international databases Web of Science and Scopus: 1. Development and Synthesis of Linguistic Models for Catalytic Cracking Unit in a Fuzzy Environment // Processes (MDPI). – 2024. – Vol. 12. – p. 1543. In addition, the research results are presented in the proceedings of an international scientific conference indexed in Scopus: 1. Evaluating machine learning-based routing algorithms on various wireless network topologies // Photonics Applications in Astronomy, Communications, Industry, and High Energy Physics Experiments (SPIE Proceedings). – 2024. – 134000Z. Structure and Volume of the Dissertation. The dissertation consists of an introduction, four chapters, a conclusion, a list of references (91 sources), and 5 appendices. The total length of the dissertation is 114 pages, including 8 tables and 30 figures. Keywords: catalytic cracking, intelligent control system, fuzzy logic, hybrid models, machine learning, uncertainty, mathematical modeling, oil refining, regression models.

Dissertation

List of Scientific Works

Consultant's Review

Foreign Consultant's Review

Conclusion of the Research Ethics Committee