✉ mojgani2[we can beat the crawlers]@illinois.edu
■ computational physics ■ data-driven modeling ■ machine learning in physical systems ■ dimensionality reduction
Model Order Reduction of convection dominated flow, Stabilization of time-varying projection based reduced order models
Aerodyamancs analysis of low-Reynolds number flapping wings
Aerodyamancs of low-Reynolds number wings with signigicant laminar to turbulent transition
A regression-type algorithm is used to discover the missing terms of a surrogate model from the observations of the corresponding high-dimensional system
A novel framework to reduce the convection dominated flows is developed. An image-registration-like approach is developed to learn an optimal manifold and the governing partial differential equations are projected and solved on this realized manifold
A new approach to stabilize time-varying ROMs is developed based on an optimal modification of the point-wise system matrices
Implemented a high-frequency arbitrage bot on GDAX (currently Coinbase Pro) Python API and deployed it on Amazon Web Services (AWS) servers
Computationally studied the aerodynamics of hover and forward flight of a low-speed flapping wing
Computationally proved the advantages of back-facing step flow on aerodynamic efficiency of an airfoil in transitional regime