«Application of Modeling and Machine Learning Methods for Optimal Planning of the Generating Equipment Composition of Pavlodar CHP-1» IRN AP09563335
Project Objective
Tasks:
— Collection and processing of telemetry data, operating mode ranges, weather data, and digitization of processes, схемs, and drawings.
— Modeling of the CHP plant and development of a machine learning algorithm for planning short-term and long-term maintenance and repair of equipment.
— Development of recommendations for improving energy efficiency and for reliability management of the CHP plant based on the Decision Tree method.
— Collection and processing of telemetry data, operating mode ranges, weather data, and digitization of processes, схемs, and drawings.
— Modeling of the CHP plant and development of a machine learning algorithm for planning short-term and long-term maintenance and repair of equipment.
— Development of recommendations for improving energy efficiency and for reliability management of the CHP plant based on the Decision Tree method.
The objective of the study is to investigate the potential for improving the energy efficiency of the CHP plant, reducing greenhouse gas emissions, and applying modeling and machine learning methods for the optimal planning of the generating equipment composition of the CHP plant in Pavlodar, Kazakhstan.
Project Relevance:
This research is important for the applied validation of big data processing methods based on machine learning for data received from an industrial facility generating electricity, heat, and industrial steam. Pavlodar CHP is unique in that it is located in the northern region of Kazakhstan and serves two different types of consumers: an aluminum plant and the population. The complexity arising from the variability of consumption profiles requires the use of modern predictive methods for planning the composition of generating equipment for a day, a week, and a year ahead. The long service life of the units and the accumulated operating hours require continuous adjustment of the annual maintenance schedule. It is necessary to take into account the dynamic demand curve for heat and electricity using forecasting methods depending on weather conditions, while the variability of heat and electricity consumption changes according to probabilistic patterns based on historical consumption data of the population.
This research is important for the applied validation of big data processing methods based on machine learning for data received from an industrial facility generating electricity, heat, and industrial steam. Pavlodar CHP is unique in that it is located in the northern region of Kazakhstan and serves two different types of consumers: an aluminum plant and the population. The complexity arising from the variability of consumption profiles requires the use of modern predictive methods for planning the composition of generating equipment for a day, a week, and a year ahead. The long service life of the units and the accumulated operating hours require continuous adjustment of the annual maintenance schedule. It is necessary to take into account the dynamic demand curve for heat and electricity using forecasting methods depending on weather conditions, while the variability of heat and electricity consumption changes according to probabilistic patterns based on historical consumption data of the population.
Results Obtained
During the implementation of the research project, a quantitative analysis of the data from CHP-1 of JSC “Aluminium of Kazakhstan” was carried out. Various unit loading levels were analyzed, and data clustering was performed. A CHP model was developed in the GAMS environment for optimal medium-term day-ahead planning of the composition of generating equipment. The numerical results of the ARIMA and Decision Tree models show that the effectiveness of the proposed approach becomes evident under conditions of high demand uncertainty, when system operating costs and fuel consumption increase due to unplanned changes in generation. The advantages of the proposed model lie in its ability to represent uncertainties when solving the stochastic long-term problem of selecting the composition of generating units. The developed algorithms can also be applied to other similar CHP plants.
During the implementation of the research project, a quantitative analysis of the data from CHP-1 of JSC “Aluminium of Kazakhstan” was carried out. Various unit loading levels were analyzed, and data clustering was performed. A CHP model was developed in the GAMS environment for optimal medium-term day-ahead planning of the composition of generating equipment. The numerical results of the ARIMA and Decision Tree models show that the effectiveness of the proposed approach becomes evident under conditions of high demand uncertainty, when system operating costs and fuel consumption increase due to unplanned changes in generation. The advantages of the proposed model lie in its ability to represent uncertainties when solving the stochastic long-term problem of selecting the composition of generating units. The developed algorithms can also be applied to other similar CHP plants.
Project Achievements
Project start and end dates: June 2021 – December 31, 2021.
Project implementation period: 7 months.
In accordance with the approved project schedule, one scientific article indexed in the Web of Science and Scopus databases was published, one article was submitted to the AUES Bulletin, and the project team members participated in an international scientific conference (Kyiv, Ukraine) and in the Young Scientists Forum (Almaty, Kazakhstan):
1. Omirgaliyev, R., Salkenov, A., Bapiyev, I., Zhakiyev, N. (2021, December). Industrial Application of Machine Learning Clustering for a Combined Heat and Power Plant: A Pavlodar Case Study. In 2021 IEEE 5th International Conference on Information and Telecommunication Technologies and Radio Electronics (UkrMiCo).
2. Sotsial Zh., Zhakiyev N., Omirgaliyev R. Application of modeling and machine learning methods for the optimal planning of CHP generating equipment composition. Proceedings of the Young Scientists Forum, Digital Kazakhstan section (September, 2021).
3. Arkhipkin O.O., Kibarin A.A., Zhakiyev N.K. Integrated approach to optimizing fuel combustion at coal-fired power plants in Kazakhstan. AUES Bulletin, No. 4, 2021 (submitted, indexed in the CQASE of the Republic of Kazakhstan).
4. Zhakiyev N., Sotsial Zh., Salkenov A., Omirgaliyev R. Set of the Data for Modeling Large-Scale Coal-Fired Combined Heat and Power Plant in Kazakhstan. Data in Brief (submitted, CiteScore > 35, Q3).
Project implementation period: 7 months.
In accordance with the approved project schedule, one scientific article indexed in the Web of Science and Scopus databases was published, one article was submitted to the AUES Bulletin, and the project team members participated in an international scientific conference (Kyiv, Ukraine) and in the Young Scientists Forum (Almaty, Kazakhstan):
1. Omirgaliyev, R., Salkenov, A., Bapiyev, I., Zhakiyev, N. (2021, December). Industrial Application of Machine Learning Clustering for a Combined Heat and Power Plant: A Pavlodar Case Study. In 2021 IEEE 5th International Conference on Information and Telecommunication Technologies and Radio Electronics (UkrMiCo).
2. Sotsial Zh., Zhakiyev N., Omirgaliyev R. Application of modeling and machine learning methods for the optimal planning of CHP generating equipment composition. Proceedings of the Young Scientists Forum, Digital Kazakhstan section (September, 2021).
3. Arkhipkin O.O., Kibarin A.A., Zhakiyev N.K. Integrated approach to optimizing fuel combustion at coal-fired power plants in Kazakhstan. AUES Bulletin, No. 4, 2021 (submitted, indexed in the CQASE of the Republic of Kazakhstan).
4. Zhakiyev N., Sotsial Zh., Salkenov A., Omirgaliyev R. Set of the Data for Modeling Large-Scale Coal-Fired Combined Heat and Power Plant in Kazakhstan. Data in Brief (submitted, CiteScore > 35, Q3).
Figure 1 – Schematic diagram of the combined heat and power plant (CHP)
Figure 2 – Big data analysis for machine learning and forecasting of CHP operation
Research Team Members
Nurhat Zhakiyev, Project Scientific Supervisor, Senior Researcher, PhD in Physics, h-index – 4 (Scopus)
https://www.scopus.com/authid/detail.uri?authorId=56043145000)
https://www.mendeley.com/authors/56043145000/
https://orcid.org/0000-0002-4904-2047
https://publons.com/researcher/D-6159-2017/
Research interests: energy modeling, combined heat and power plants, physics. Authored 15 scientific publications. Project supervisor for grant-funded project in 2021.
Key publications related to the project:
[1] Kopanos G., Murele O.C., Silvente J., Zhakiyev N., Akhmetbekov Y., Tutkushev D. (2018). Efficient planning of energy production and maintenance of large-scale combined heat and power plants. Energy Conversion and Management,169,390-403 (Q1) https://doi.org/10.1016/j.enconman.2018.05.022
[2] Zhakiyev, N., Akhmetbekov, Y., Silvente, J., & Kopanos, G. M. (2017). Optimal energy dispatch and maintenance of an industrial coal-fired combined heat and power plant in Kazakhstan. Energy Procedia, 142, 2485-2490. https://doi.org/10.1016/j.egypro.2017.12.187
[3] Zhakiyev, N., & Otarov, R. (2017). Scheduling and planning for optimal operations of power plants using a unit commitment approach. In Sustainable Energy in Kazakhstan: Moving to Cleaner Energy in a Resource-Rich Country (pp. 109-115). Taylor and Francis. https://doi.org/10.4324/9781315267302
[4] Omirgaliyev, R., Salkenov, A, Bapiyev, I, Zhakiyev N. (2021, December). Industrial Application of Machine Learning Clustering for a Combined Heat and Power Plant: A Pavlodar Case Study. In 2021 IEEE 5th International Conference on Information and Telecommunication Technologies and Radio Electronics (UkrMiCo).
[5] Социал Ж., Жакив Н., Омиргалиев Р., Применение методов моделирования и машинного обучения для оптимального планирования состава генерирующего оборудования ТЭЦ, Сборник тезисов Форума молодых ученых в секции Цифровой Казахстан (сентябрь, 2021)
[6] Zhakiyev N., Sotsial Zh., Salkenov A., Omirgaliyev R. Set of the Data for Modeling large-scale Coal-Fired Combined Heat and Power Plant in Kazakhstan
Data in Brief, (Submitted, Dec, 2021, CiteScore>35, Q3)
https://www.scopus.com/authid/detail.uri?authorId=56043145000)
https://www.mendeley.com/authors/56043145000/
https://orcid.org/0000-0002-4904-2047
https://publons.com/researcher/D-6159-2017/
Research interests: energy modeling, combined heat and power plants, physics. Authored 15 scientific publications. Project supervisor for grant-funded project in 2021.
Key publications related to the project:
[1] Kopanos G., Murele O.C., Silvente J., Zhakiyev N., Akhmetbekov Y., Tutkushev D. (2018). Efficient planning of energy production and maintenance of large-scale combined heat and power plants. Energy Conversion and Management,169,390-403 (Q1) https://doi.org/10.1016/j.enconman.2018.05.022
[2] Zhakiyev, N., Akhmetbekov, Y., Silvente, J., & Kopanos, G. M. (2017). Optimal energy dispatch and maintenance of an industrial coal-fired combined heat and power plant in Kazakhstan. Energy Procedia, 142, 2485-2490. https://doi.org/10.1016/j.egypro.2017.12.187
[3] Zhakiyev, N., & Otarov, R. (2017). Scheduling and planning for optimal operations of power plants using a unit commitment approach. In Sustainable Energy in Kazakhstan: Moving to Cleaner Energy in a Resource-Rich Country (pp. 109-115). Taylor and Francis. https://doi.org/10.4324/9781315267302
[4] Omirgaliyev, R., Salkenov, A, Bapiyev, I, Zhakiyev N. (2021, December). Industrial Application of Machine Learning Clustering for a Combined Heat and Power Plant: A Pavlodar Case Study. In 2021 IEEE 5th International Conference on Information and Telecommunication Technologies and Radio Electronics (UkrMiCo).
[5] Социал Ж., Жакив Н., Омиргалиев Р., Применение методов моделирования и машинного обучения для оптимального планирования состава генерирующего оборудования ТЭЦ, Сборник тезисов Форума молодых ученых в секции Цифровой Казахстан (сентябрь, 2021)
[6] Zhakiyev N., Sotsial Zh., Salkenov A., Omirgaliyev R. Set of the Data for Modeling large-scale Coal-Fired Combined Heat and Power Plant in Kazakhstan
Data in Brief, (Submitted, Dec, 2021, CiteScore>35, Q3)
Ruslan Omirgaliyev, Junior Researcher, holds a Master’s degree in Electrical Engineering.
Research interests: physics, mathematics, programming, data analytics.
Key publications related to the project area:
1. Omirgaliyev, R., Salkenov, A., Bapiyev, I., Zhakiyev, N. (2021, December). Industrial Application of Machine Learning Clustering for a Combined Heat and Power Plant: A Pavlodar Case Study. In 2021 IEEE 5th International Conference on Information and Telecommunication Technologies and Radio Electronics (UkrMiCo).
2. Sotsial, Zh., Zhakiyev, N., Omirgaliyev, R. Application of Modeling and Machine Learning Methods for Optimal Planning of Generating Equipment Composition at the CHP Plant. Collection of abstracts of the Young Scientists Forum, Digital Kazakhstan section (September 2021).
Research interests: physics, mathematics, programming, data analytics.
Key publications related to the project area:
1. Omirgaliyev, R., Salkenov, A., Bapiyev, I., Zhakiyev, N. (2021, December). Industrial Application of Machine Learning Clustering for a Combined Heat and Power Plant: A Pavlodar Case Study. In 2021 IEEE 5th International Conference on Information and Telecommunication Technologies and Radio Electronics (UkrMiCo).
2. Sotsial, Zh., Zhakiyev, N., Omirgaliyev, R. Application of Modeling and Machine Learning Methods for Optimal Planning of Generating Equipment Composition at the CHP Plant. Collection of abstracts of the Young Scientists Forum, Digital Kazakhstan section (September 2021).
Aldiyar Kanatovich Salkenov, Junior Researcher, holds a Master’s degree in Information Technology.
Field of research interests: web development, data analytics.
Key publications related to the project area:
1. Omirgaliyev, R., Salkenov, A., Bapiyev, I., Zhakiyev, N. (2021, December). Industrial Application of Machine Learning Clustering for a Combined Heat and Power Plant: A Pavlodar Case Study. In 2021 IEEE 5th International Conference on Information and Telecommunication Technologies and Radio Electronics (UkrMiCo, 29.11–03.12, 2021). (Submitted, indexed in IEEE/Scopus/WoS).
Field of research interests: web development, data analytics.
Key publications related to the project area:
1. Omirgaliyev, R., Salkenov, A., Bapiyev, I., Zhakiyev, N. (2021, December). Industrial Application of Machine Learning Clustering for a Combined Heat and Power Plant: A Pavlodar Case Study. In 2021 IEEE 5th International Conference on Information and Telecommunication Technologies and Radio Electronics (UkrMiCo, 29.11–03.12, 2021). (Submitted, indexed in IEEE/Scopus/WoS).
Zhuldyz Zheniskyzy Sotsial, Junior Researcher, holds a Master’s degree in Mechanical and Aerospace Engineering.
Field of research interests: mathematics, programming.
Key publications related to the project area:
1. Sotsial Zh., Zhakiyev N., Omirgaliyev R. Application of modeling and machine learning methods for the optimal planning of CHP generating equipment composition. Proceedings of the Young Scientists Forum, Digital Kazakhstan section (September, 2021).
2. Zhakiyev N., Sotsial Zh., Salkenov A., Omirgaliyev R. Set of the Data for Modeling Large-Scale Coal-Fired Combined Heat and Power Plant in Kazakhstan.
Data in Brief (submitted, December 2021, CiteScore > 35, Q3).
Field of research interests: mathematics, programming.
Key publications related to the project area:
1. Sotsial Zh., Zhakiyev N., Omirgaliyev R. Application of modeling and machine learning methods for the optimal planning of CHP generating equipment composition. Proceedings of the Young Scientists Forum, Digital Kazakhstan section (September, 2021).
2. Zhakiyev N., Sotsial Zh., Salkenov A., Omirgaliyev R. Set of the Data for Modeling Large-Scale Coal-Fired Combined Heat and Power Plant in Kazakhstan.
Data in Brief (submitted, December 2021, CiteScore > 35, Q3).