Efficient reduced models and a posteriori error estimation for parametrized dynamical systems by offline/online decomposition

摘要: We address the problem of model order reduction (MOR) parametrized dynamical systems. Motivated by reduced basis (RB) methods for partial differential equations, we show that some characteristic components can be transferred to linear assume an affine parameter dependence system components, which allows offline/online decomposition and is efficient simulation. Additionally, error control possible a posteriori estimators state vector output vector, based on residual analysis primal-dual techniques. Experiments demonstrate applicability systems, reliability runtime gain technique. The estimation technique straightforwardly applied all traditional projection-based techniques non-parametric parametric such as reduction, balanced truncation, mom...