You are here

Methods for Data-driven Modeling of Unsteady Fluid Flows

Peter J. Baddoo, MIT

Benjamin Herrmann, University of Washington

Advances in our capability to find patterns and identify dynamics in unsteady fluid flows will improve our ability to predict, control, and understand a range of systems relevant to science and engineering. However, these dynamical systems are typically difficult to model because they are nonlinear, nonnormal, and high-dimensional. Fortunately, techniques from statistics, convex optimization, and machine learning can be leveraged to find and describe the evolution of coherent structures that dominate dynamic activity and govern engineering quantities of interest. This symposium will bring together recent research efforts on the development of novel data-driven methods and their application to pressing challenges in aeronautics, biomedicine and energy conversion.