Researcher
E-mail: l.haoyan.gm@gmail.com
Location: Avenue F. D. Roosevelt, 50 bte CP165/41
Career Summary:
Haoyan Li began his PhD in October 2023 as a member of the Marie Skłodowska-Curie project ModelAir. His research focuses on the development of reduced-order models (ROMs) for urban flows in the Brussels district, along with the uncertainty quantification of turbulence models.
He obtained his Master’s degree in Mechanics and Mathematical Modelling from Novosibirsk State University, Russia, graduating with the highest distinction. During his Master’s studies, he worked on enhancing turbulence models for the accurate representation of secondary flows.
In his PhD work, Haoyan develops ROMs for turbulent and urban flows using both linear approaches — such as Principal Component Analysis (PCA) — and non-linear approaches — such as autoencoders. He proposed a domain-decomposition ROM based on clustering techniques using linear methods and developed a novel β-variational autoencoder capable of accurately predicting turbulent flows. In addition, his work addresses the uncertainty quantification of turbulence models and the development of non-linear eddy viscosity models tailored for urban flow simulations.
Main Research Interests:
- Reduced-Order Model
- Uncertainty Quantification
- Turbulence Modelling
- Turbulent Flows
- Urban Flows
- Machine Learning
- Deep Learning
- CFD
- Fluid Dynamics
- Simulations
Involvement in Projects and Grants: