Doklady Tomskogo gosudarstvennogo universiteta sistem upravleniya i radioelektroniki / Archive / Issue of Journal № 3, Vol. 28, 2025 / Machine learning models for forecasting planned levels of mineral production

Machine learning models for forecasting planned levels of mineral production

DOI: 10.21293/1818-0442-2025-28-3-38-44

Download article in PDF format

JATS xml

Abstract: The oil and gas industry of the Russian Federation, a powerful driver of the country's economic development, directly depends on the speed of commissioning new mining fields. The decline in the accuracy of hydrocarbon production forecasts by Russian companies is a consequence of the deteriorating quality of their resource base. This study evaluates the effectiveness of ma-chine-learning-based predictive models for forecasting hydro-carbon production volumes. A method for training predictive neural network models is described, incorporating a dataset of geological, geophysical and design indicators for the develop-ment of oil and gas fields. Missing geological and geophysical data were reconstructed using various augmentation methods.

Keywords: GRU, LSTM, MLP, CNN1D, data augmentation, hydrocarbon feedstock, production facility, data normalization

For citation:
Dergachev A. O., Borovskoy I. G., Kruchinin V. V. Machine learning models for forecasting planned levels of mineral production. Doklady Tomskogo gosudarstvennogo universiteta sistem upravleniya i radioelektroniki, 2025, vol. 28, no. 3, pp. 38–44. DOI: 10.21293/1818-0442-2025-28-3-38-44

Authors and copyright holders:

Dergachev A. O. , Tomsk State University of Control Systems and Radioelectronics (Tomsk, Russia)
Borovskoy I. G. , Tomsk State University of Control Systems and Radioelectronics (Tomsk, Russia)
Kruchinin V. V. , Tomsk State University of Control Systems and Radioelectronics (Tomsk, Russia)

1. Chto takoe neyroset I kak ona rabotaet [What are neural networks and how they work]. Available at: https://sky.pro/media/neyronnye-seti/ (Accessed 21.02.2025). (in Russ.).
2. Shcherbakova K.O. [The problem of high water cut in the products of horizontal wells] Proceedings of higher educational stablishments. Geology and Exploration, 2022, vol. 64, no. 6, pp. 29–38. DOI: https://www.geology-mgri.ru/jour/article/view/844 (in Russ.).
3. Kachnov S.A. Obzor metodov augmentatsii tekstovyh dannyh [Overview of Text Data Augmentation Methods] Sbornik nauchnyh statei 6-i Mezhdunarodnoi nauchnoi konferentsii perspektivnyh razrabotok molodyh uchenyh [Collection of scientific articles from the 6th International Scientific Conference on Advanced Developments of Young Scientists]. Kursk, South-West State University, 2021. Vol. 4, pp. 55–58 (in Russ.).
4. Rekurrentnaya neironnaya set (RNN): vidy, obuchenie, primery [Recurrent neural network (RNN): types training, examples]. Available at: https://neurohive.io/ru/osnovy-data-cience/rekurrentnye-nejronnye-seti / (Accessed: 11.03.2025) (in Russ.).
5. Back propagation – algoritm obucheniya po metodu obratnogo rasprostraneniya [Back propagation – learning algorithm by the method of back propagation]. Available at: https://proproprogs.ru/neural_network/back-propa gation-algoritm-obucheniya-po-metodu-obratnogo-rasprostraneniya (Accessed: 12.03.2025) (in Russ.).
6. Biblioteka glubokogo obucheniya Tensorflow [Tensorflow deep learning Library]. Available at: https://habr.com/ru/companies/ods/articles/324898/ (Accessed: 25.01.2025) (in Russ.).
7. Parametry, vliyayushchie na effektivnost' neironnoi seti [Parameters affecting the efficiency of a neural network]. Available at: http://www.100byte.ru/python/factors/factors.html (Accessed: 17.03.2025) (in Russ.).
8. Borba s pereobucheniem neironnyh setei: prichiny, vliyanie i metody predotvrashcheniya [Combating neural network overfitting: causes, impact, and prevention methods]. Available at: https://na-journal.ru/5-2024-informacionnyetekhnologii/12045-borba-s-pereobucheniem-neironnyh-seteiprichiny-vliyanie-i-metody-predotvrashchen iya/ (Accessed 29.03.2025) (in Russ.).
9. Nikolenko S.I., Kadurin A.A., Arkhangelskaya E.O. Glubokoe obuchenie. Pogruzhenie v mir neironnyh setei [Deep learning. Immersion in the world of neural networks]. St. Petersburg, Piter, 2018, 480 p. (in Russ.).
10. Goodfellow Ya., Bendjio I., Courville A. Glubokoe obuchenie [Deep learning]. 2nd ed., corr., Moscow, DMK Press, 2018, 652 p. (in Russ.).
11. Tarik R. Sozdayom neironnuyu set' [Creating a neural network]. St. Petersburg, Alfa-book LLC, 2017, 272 p. (in Russ.).
12. Scholle F. Glubokoe obuchenie na Python [Deep learning in Python]. 2nd ed., St. Petersburg, Piter, 2023, 576 p. (in Russ.).
13. Murphy K.P. Veroyatnostnoe mashinnoe obuchenie [Probabilistic machine learning]. Moscow, DMK Press, 2022, 940 p. (in Russ.).
14. Nikolenko S.I. Mashinnoe obuchenie [Machine learning]. St. Petersburg, Peter, 2025, 608 p. (in Russ.).
15. Howard D., Gugger S. Glubokoe obuchenie [Deep learning]. St. Petersburg, Peter, 2022, 624 p. (in Russ.).