Development of Control Algorithms for Multiphase Flow in Porous Media Using Neural Operators

Авторы: Sirota Daniil D., Dushin Sergey E.
Сборник: 2025 VI International Conference on Control in Technical Systems (CTS)
Год издания: 2025
Место издания: IEEE
Первая страница: 7
Последняя страница: 10
DOI: 10.1109/cts67336.2025.11196690
Аннотация: Numerical simulation of fluid flow in porous media is associated with high computational complexity, which significantly limits the applicability of optimal control algorithms to such processes. At the same time, effective control of flow processes is of great practical importance in fields such as oil and gas production, operation of underground gas storage (UGS) facilities, and geological carbon dioxide sequestration. This study proposes an approach that overcomes the computational limitations of traditional numerical methods by leveraging modern deep learning techniques. The primary objective is to develop an optimal control algorithm based on the approximation of nonlinear flow equations using neural operators. The trained models enable a substantial reduction in computational cost when solving control problems. The paper presents validation results for the proposed approach on a model problem of CO2 injection into a reservoir, demonstrating its computational efficiency and potential for practical application to real-world systems.
Добавил в систему: Сирота Даниил Дмитриевич

	ИСТИНА	Войти в систему Регистрация
	ИСТИНА ПсковГУ
	Главная Поиск Статистика О проекте Помощь

ИСТИНА