Principles of creating a digital twin prototype for the process of alkylation of benzene with propylene based on a neural network
Objectives. To identify the principles of creating digital twins of an operating technological unit along the example of the process of liquid-phase alkylation of benzene with propylene, and to establish the sequence of stages of formation of a digital twin, which can be applied to optimize oil and...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | Russian |
| Published: |
MIREA - Russian Technological University
2023-11-01
|
| Series: | Тонкие химические технологии |
| Subjects: | |
| Online Access: | https://www.finechem-mirea.ru/jour/article/view/2001 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850028779035951104 |
|---|---|
| author | K. G. Kichatov T. R. Prosochkina I. S. Vorobyova |
| author_facet | K. G. Kichatov T. R. Prosochkina I. S. Vorobyova |
| author_sort | K. G. Kichatov |
| collection | DOAJ |
| description | Objectives. To identify the principles of creating digital twins of an operating technological unit along the example of the process of liquid-phase alkylation of benzene with propylene, and to establish the sequence of stages of formation of a digital twin, which can be applied to optimize oil and gas chemical production.Methods. The chemical and technological system consisting of reactor, mixer, heat exchangers, separator, rectification columns, and pump is considered as a complex high-level system. Data was acquired in order to describe the functioning of the isopropylbenzene production unit. The main parameters of the process were calculated by simulation modeling using UniSim® Design software. A neural network model was developed and trained. The influence of various factors of the reaction process of alkylation, separation of reaction products, and evaluation of economic factors providing market interest of the industrial process was also considered. The adequacy of calculations was determined by statistics methods. A microcontroller prototype of the process was created.Results. A predictive neural network model and its creation algorithm for the process of benzene alkylation was developed. This model can be loaded into a microcontroller to allow for real-time determination of the economic efficiency of plant operation and automated optimization depending on the following factors: composition of incoming raw materials; the technological mode of the plant; the temperature mode of the process; and the pressure in the reactor.Conclusions. The model of a complex chemicotechnological system of cumene production, created and calibrated on the basis of long-term industrial data and the results of calculations of the output parameters, enables the parameters of the technological process of alkylation to be calculated (yield of reaction products, energy costs, conditional profit at the output of finished products). During the development of a hardware-software prototype, adapted to the operation of the real plant, the principles and stages of creating a digital twin of the operating systems of chemical technology industries were identified and formulated. |
| format | Article |
| id | doaj-art-b637eb8e55644fb29887c19e44cbf631 |
| institution | DOAJ |
| issn | 2410-6593 2686-7575 |
| language | Russian |
| publishDate | 2023-11-01 |
| publisher | MIREA - Russian Technological University |
| record_format | Article |
| series | Тонкие химические технологии |
| spelling | doaj-art-b637eb8e55644fb29887c19e44cbf6312025-08-20T02:59:43ZrusMIREA - Russian Technological UniversityТонкие химические технологии2410-65932686-75752023-11-0118548249710.32362/2410-6593-2023-18-5-482-4971735Principles of creating a digital twin prototype for the process of alkylation of benzene with propylene based on a neural networkK. G. Kichatov0T. R. Prosochkina1I. S. Vorobyova2Technological Faculty, Ufa State Petroleum Technological UniversityTechnological Faculty, Ufa State Petroleum Technological UniversityTechnological Faculty, Ufa State Petroleum Technological UniversityObjectives. To identify the principles of creating digital twins of an operating technological unit along the example of the process of liquid-phase alkylation of benzene with propylene, and to establish the sequence of stages of formation of a digital twin, which can be applied to optimize oil and gas chemical production.Methods. The chemical and technological system consisting of reactor, mixer, heat exchangers, separator, rectification columns, and pump is considered as a complex high-level system. Data was acquired in order to describe the functioning of the isopropylbenzene production unit. The main parameters of the process were calculated by simulation modeling using UniSim® Design software. A neural network model was developed and trained. The influence of various factors of the reaction process of alkylation, separation of reaction products, and evaluation of economic factors providing market interest of the industrial process was also considered. The adequacy of calculations was determined by statistics methods. A microcontroller prototype of the process was created.Results. A predictive neural network model and its creation algorithm for the process of benzene alkylation was developed. This model can be loaded into a microcontroller to allow for real-time determination of the economic efficiency of plant operation and automated optimization depending on the following factors: composition of incoming raw materials; the technological mode of the plant; the temperature mode of the process; and the pressure in the reactor.Conclusions. The model of a complex chemicotechnological system of cumene production, created and calibrated on the basis of long-term industrial data and the results of calculations of the output parameters, enables the parameters of the technological process of alkylation to be calculated (yield of reaction products, energy costs, conditional profit at the output of finished products). During the development of a hardware-software prototype, adapted to the operation of the real plant, the principles and stages of creating a digital twin of the operating systems of chemical technology industries were identified and formulated.https://www.finechem-mirea.ru/jour/article/view/2001digital twincumeneindustrial plantneural networksmachine learningesp8266 |
| spellingShingle | K. G. Kichatov T. R. Prosochkina I. S. Vorobyova Principles of creating a digital twin prototype for the process of alkylation of benzene with propylene based on a neural network Тонкие химические технологии digital twin cumene industrial plant neural networks machine learning esp8266 |
| title | Principles of creating a digital twin prototype for the process of alkylation of benzene with propylene based on a neural network |
| title_full | Principles of creating a digital twin prototype for the process of alkylation of benzene with propylene based on a neural network |
| title_fullStr | Principles of creating a digital twin prototype for the process of alkylation of benzene with propylene based on a neural network |
| title_full_unstemmed | Principles of creating a digital twin prototype for the process of alkylation of benzene with propylene based on a neural network |
| title_short | Principles of creating a digital twin prototype for the process of alkylation of benzene with propylene based on a neural network |
| title_sort | principles of creating a digital twin prototype for the process of alkylation of benzene with propylene based on a neural network |
| topic | digital twin cumene industrial plant neural networks machine learning esp8266 |
| url | https://www.finechem-mirea.ru/jour/article/view/2001 |
| work_keys_str_mv | AT kgkichatov principlesofcreatingadigitaltwinprototypefortheprocessofalkylationofbenzenewithpropylenebasedonaneuralnetwork AT trprosochkina principlesofcreatingadigitaltwinprototypefortheprocessofalkylationofbenzenewithpropylenebasedonaneuralnetwork AT isvorobyova principlesofcreatingadigitaltwinprototypefortheprocessofalkylationofbenzenewithpropylenebasedonaneuralnetwork |