MS7-02 - Recent trends in model order reduction for fluid dynamics
The numerical simulation of parametrized fluid dynamic problems require the resolution of multiple high dimensional systems of nonlinear equations which usually require enormous computational resources. This fact is particularly evident in the case of complex geometries and turbulent flows . Reduced order models  offer an attractive methodology in order to reduce the computational cost while retaining an acceptable level of accuracy. This minisymposium is devoted to recent trends for the computational reduction of fluid dynamics problems using both intrusive, non-intrusive and hybrid approaches. Some of the considered topics are reduced basis methods , machine learning approaches, structure-preserving reduced order models, Gaussian process regression, dynamic mode decomposition. Possible applications include, but are not limited to, uncertainty quantification, turbulent flows, inverse problems, real-time control, shape optimization, etc.