In this project, students will gain hands-on experience with physics-informed machine learning (PINNs), learning how to integrate physical laws with AI to improve fluid flow modeling in porous media. They will develop skills in machine learning algorithms, fluid dynamics, and computational tools like Python and TensorFlow. Students will also enhance their data-handling, problem-solving, and critical thinking abilities while working collaboratively across disciplines. This project will equip them with valuable skills applicable to fields like engineering, energy, and environmental science, preparing them for both academic and industry roles
Project Info
Physics-informed machine learning for fluid flow in porous media
Pejman Tahmasebi
tahmasebi@mines.edu
Project Goals and Description:
This project focuses on the integration of machine learning techniques with physics-based models to better understand and predict fluid flow in porous media. By leveraging physics-informed neural networks (PINNs), we aim to combine the predictive power of machine learning with the governing equations of fluid dynamics, such as the Navier-Stokes equations and Darcy's law, to create more accurate models for multiphase fluid flow in complex porous structures. The goal is to enhance our ability to simulate and analyze fluid behavior in applications ranging from oil and gas recovery to environmental remediation and material science.
The goals of this project are compelling because they address critical challenges in a variety of fields that rely on accurate fluid flow modeling. Traditional numerical methods for solving fluid flow problems in porous media can be computationally expensive and may not efficiently handle the complexity of real-world systems. By incorporating machine learning into the modeling process, this project aims to develop faster, more efficient solutions that maintain physical accuracy. The use of physics-informed machine learning not only improves model performance but also opens new avenues for innovation in industries such as energy, environmental engineering, and infrastructure development. Furthermore, this project offers students the opportunity to work at the cutting edge of AI and engineering, equipping them with skills that are highly relevant to both academic research and industry.
More Information:
Grand Challenge: Develop carbon sequestration methods.
Primary Contacts:
Tahmasebi, Pejman
Student Preparation
Qualifications
Coding in python
TIME COMMITMENT (HRS/WK)
5
SKILLS/TECHNIQUES GAINED
MENTORING PLAN
During the course of this project, I plan to mentor the students through regular meetings, where we will discuss their progress, address any challenges, and ensure they stay on track. These meetings will provide a space to go over issues related to code development, helping students troubleshoot problems and explore potential solutions. I'll guide them through debugging, optimizing their code, and applying machine learning techniques effectively. By providing continuous feedback, I aim to help them not only overcome technical challenges but also develop critical thinking skills for problem-solving. Additionally, I'll encourage collaboration and open communication to foster a supportive learning environment throughout the project.
Preferred Student Status
Sophomore
Junior
Senior