Automation of Neural Network Forecasting of Emergency Situations in the Management Tasks of Oil and Gas Industry

   This study addresses the automation of predicting emergency situations in control systems through artificial-intelligence methods. As a case example, well-operation control is examined: the prediction of emergencies is carried out by processing large volumes of data comprising sensor readings from the hoisting-unit telemetry system and routine event logs from the well. To handle these data sets, a recurrent neural network is proposed. A mathematical model is constructed, on the basis of which a structural model of the prototype prediction program integrating data and knowledge is developed.
   A conceptual architectural scheme of the neural-network-based program for automating emergency prediction in control systems is presented, together with an information model and a description of the program’s operating principles. The program is implemented in Python using the Pandas and PyTorch libraries. The paper reports the resulting performance metrics, which also enable related tasks such as linking emergency situations to specific crews and repair types. Neuralnetwork varieties suited to particular subtasks are discussed: recurrent neural networks (RNN), convolutional neural networks (CNN), and long short-term memory networks (LSTM). Developing a neural-network program for automated emergency prediction represents an important tool for improving safety, reducing risks, and optimizing production processes in the oiland gas industry.

1. Zuev S. V. Statistical online learning in recurrent and direct propagation of quantum neural networks, Reports of the Russian Academy of Sciences. Mathematics, Informatics, Control Processes, 2023, vol. 514, no. 2, pp. 177—186 (in Russian).

2. Shchelchkov Yu. E. Analysis of the oil and gas industry of the Russian Federation, NovaInfo, 2023, no. 138, pp. 51—53 (in Russian).

3. Kostyuchek M. I., Makarenko A. V. Application of the convolutional deep neural networks for the solution of some tasks of the trajectory data analysis, Journal of radio electronics, 2021, no. 11, pp. 1—25 (in Russian).

4. Skripachev V. O., Guida M. V., Guida N. V., Zhukov A. O. Features of convolutional neural networks work, International Journal of Open Information Technologies, 2022, vol. 10, no. 12, pp. 53—61 (in Russian).

5. Meizer M. V., Severianova E. D., Mokshin V. V. Example of time series forecasting using recurrent neural network LSTM, Young Scientist, 2022, no. 9 (404), pp. 13—15, available at: https://moluch.ru/archive/404/89260/ (in Russian)

6. Zhiganov V., Ivanov Yu.S., Grabar D. M. Investigation of neural network algorithms for human movement prediction on the basis of LSTM and Transformers, Reports of the Russian Academy of Sciences. Mathematics, Informatics, Control Processes, 2023, vol. 514, no. 2, pp. 364—374 (in Russian).

7. Cholakov R., Kolev T. Transformers predicting the future. Applying attention in next-frame and time series forecasting, arXiv:2108.08224v1 [cs.LG], 18 Aug 2021.

8. Antonov V. V., Kromina L. A., Rodionova L. E., Fakhrullina A. R., Baimurzina L. I., Palchevsky E. V., Rodionov E. A. The concept of forming the intellectual control systems of the power supply of the city networks, Мekhatronica, Avtomatizatsiya, Upravlenie, 2023, vol. 24, no. 4, pp. 190—198. DOI: 10.17587/mau.24.190-198 (in Russian).

9. Palchevsky E. V., Antonov V. V., Kromina L. E., Rodionova L. E., Fakhrunina A. R. Intelligent Forecasting of Electricity Consumption in Managing Energy Enterprises in Order to Carry out Energy-Saving Measures, Мekhatronica, Avtomatizatsiya, Upravlenie, 2023, vol. 24, no. 6, pp. 307—316, DOI: 10.17587/mau.24.307-316

10. Tikhonov D. V., Ermolenko M. O. Modern technologies of applied programming and data processing, Yaroslavl, LLC "PKF ‘SOYUZ-PRESS’, 2024, 232 p. (in Russian).

11. Sukhov E. A., Ryabykin A. S. Neural network methods in the task of sentiment analysis, Problems of the digital signal processing application, 2022, no. 12, pp. 41—57 (in Russian).

12. Rapakov G. G., Gorbunov V. A., Dianov S. V., Elizarova L. V. Research LSTM-neural network approach in modeling of time series, Vestnik of Cherepovets State University, 2023, no. 3 (114), pp. 47—54 (in Russian).

13. Tormozov V. S. Adaptation of the LSTM neural network model for solving the complex problem of pattern recognition, Program products and systems, 2021, no. 1, pp. 151—156 (in Russian).

14. Abdulnagimov A. I., Antonov V. V., Chepaykin A. S., Palchevsky E. V., Nasyrov N. T. Neural Network Simulator of Gas Turbine Engines for Prototyping and Debugging of Control Systems on Unsteady Modes, Мekhatronica, Avtomatizatsiya, Upravlenie, 2025, vol. 26, no. 1, pp. 22—27. DOI: 10.17587/mau.26.22-27

15. Fedorov I. E., Gaar M. P., Garifullin A. E. Analysis and application of ELU activation functions of derivative functions, Vestnik Novaya Era, 2024, pp. 204—217 (in Russian).

16. Palchevsky E., Antonov V., Filimonov N., Rodionova L. E. [et al.] Devel-opment of an intelligent system for early forecasting and modelling of flood situa-tion on the example of the Republic of Bashkortostan using a proprietary ma-chine and deep learning library, Journal of Hydrology, 2024, vol. 633, pp. 130978.

17. Anisimov A. A., Sorokovnin M. E., Tararykin S. V. Neural network identification and tuning of the mechatronic systems with the state regulators, Мekhatronica, Avtomatizatsiya, Upravlenie, 2025, vol. 26, no. 2, pp. 65—76 (in Russian). DOI: 10.17587/mau.26.65-76.

18. Konstantinov A. V., Utkin L. V. New computationally simple method for realization of the neural networks with rigid restrictions on the output data, Reports of the Russian Academy of Sciences. Mathematics, Informatics, Control Processes, 2023, vol. 514, no. 2, pp. 80—90 (in Russian).