DDPS | Hybrid reduced order models
“Hybrid reduced order models: from exploiting physical principles to novel machine learning approaches”
DDPS Talk date: January 31st, 2024
Speaker: Soledad Le Clainche (Universidad Politecnica de Madrid, https://sites.google.com/view/soledad...)
Description: In the fight against climate change, humanity faces one of its most significant challenges. Tackling this issue requires the exploration of various approaches and the promotion of innovative technologies aimed at reducing atmospheric pollution. Fluid mechanics plays a vital role, offering a wide range of applications—from improving the efficiency of combustion systems to developing strategies for controlling urban air pollution and enhancing aircraft design for better fuel efficiency.
This research focuses on developing reduced-order models (ROMs) that are rooted in fundamental physical principles. These models are constructed using modal decomposition techniques such as singular value decomposition (SVD) and higher-order dynamic mode decomposition (HODMD). Additionally, machine learning methods like neural networks are integrated into the modeling process. By blending these traditional and modern approaches, the study aims to develop accurate and efficient predictive ROMs that can address a variety of challenges in fluid dynamics.
The methodologies developed in this study have a broad array of applications. In energy systems, they can enhance the performance of combustion processes, reducing harmful emissions. In urban environments, the models can be used to create more effective strategies for managing air pollution. Additionally, these tools are highly relevant in aerodynamics, where they can be employed to refine aircraft designs, resulting in greater fuel efficiency and lower carbon footprints. In the domain of flow control, the new techniques introduced here could lead to advanced methods for managing fluid behavior in various industrial processes, contributing further to the global effort to reduce environmental impact.
Bio: Dr. Soledad Le Clainche, is Professor of Applied Mathematics at the School of Aeronautics of the Universidad Politécnica de Madrid (UPM). In December 2013 she finished her PhD at the same University, in the Dept. of Fluid Dynamics and Aerospace Propulsion. Her main lines of research focus on computational fluid dynamics, data analysis, machine learning and the development and application of predictive reduced order models based on physical principles. In this line, she is PI of several national and EU-funded projects whose main objective is to develop new strategies to reduce air pollution in cities, to develop more efficient combustion systems and aerodynamic designs, and to advance in the field of personalized medicine.
DDPS webinar: https://www.librom.net/ddps.html
💻 LLNL News: https://www.llnl.gov/news
📲 Instagram: / livermore_lab
🤳 Facebook: / livermore.lab
🐤 Twitter: / livermore_lab
About LLNL: Lawrence Livermore National Laboratory has a mission of strengthening the United States’ security through development and application of world-class science and technology to: 1) enhance the nation’s defense, 2) reduce the global threat from terrorism and weapons of mass destruction, and 3) respond with vision, quality, integrity and technical excellence to scientific issues of national importance. Learn more about LLNL: https://www.llnl.gov/.
IM release number is: LLNL-VIDEO-872374