• Corpus ID: 1333656

A numerically stable a posteriori error estimator for reduced basis approximations of elliptic equations

@article{Buhr2014ANS,
  title={A numerically stable a posteriori error estimator for reduced basis approximations of elliptic equations},
  author={Andreas Buhr and Christian Engwer and Mario Ohlberger and Stephan Rave},
  journal={arXiv: Numerical Analysis},
  year={2014}
}
AbstractThe Reduced Basis (RB) method is a well established method for the model order reductionof problems formulated as parametrized partial di erential equations. One crucial requirementfor the application of RB schemes is the availability of an a posteriori error estimator toreliably estimate the error introduced by the reduction process. However, straightforwardimplementations of standard residual based estimators show poor numerical stability, renderingthem unusable if high accuracy is… 

Figures from this paper

Multilevel a posteriori error estimator for greedy reduced basis algorithms
TLDR
A new kind of the estimate is introduced, which is based on a multilevel splitting of a discretized solution space, and it is compared with a standard estimate based on bounds to coercivity and continuity constants, which can be obtained for both methods.
A hierarchical a posteriori error estimator for the Reduced Basis Method
TLDR
This work considers the Reduced Basis Method in a Petrov-Galerkin framework, and proposes and analyzes a hierarchical a posteriori error estimator which evaluates the difference of two reduced approximations of different accuracy.
A robust error estimator and a residual-free error indicator for reduced basis methods
Towards Automatic and Reliable Localized Model Order Reduction
TLDR
This thesis investigates different aspects of localized model order reduction and proposes various improvements, and a simulation methodology named ArbiLoMod, comprising a localized training, a localized a posteriori error estimator and an enrichment procedure is proposed.
Localized model reduction for parameterized problems
TLDR
In detail, it is shown how optimal local approximation spaces can be constructed and approximated by random sampling and an overview of possible conforming and non-conforming couplings of the local spaces is provided.
Randomized linear algebra for model reduction. Part I: Galerkin methods and error estimation
TLDR
A probabilistic way for reducing the cost of classical projection-based model order reduction methods for parameter-dependent equations by considering a random sketch of the full order model, which is a set of low-dimensional random projections of large matrices and vectors involved in theFull order model.
RANDOMIZED LOCAL MODEL ORDER REDUCTION \ast
TLDR
An adaptive randomized algorithm based on methods from randomized linear algebra is proposed which constructs a local reduced space approximating the range of the transfer operator and thus the optimal local approximation spaces.
Certified Reduced Basis Approximation for the Coupling of Viscous and Inviscid Parametrized Flow Models
TLDR
The main contributions of this paper are the combination of an offline-online splitting with the domain decomposition approach, reducing both offline and online computational loads and a new kernel interpolation method for the approximation of the stability factor in the online evaluation of the error estimate.
Randomized Local Model Order Reduction
TLDR
An adaptive randomized algorithm based on methods from randomized linear algebra is proposed, which constructs a local reduced space approximating the range of the transfer operator and thus the optimal local approximation spaces.
Model reduction for fractional elliptic problems using Kato's formula
TLDR
This work proposes a novel numerical algorithm utilizing model reduction for computing solutions to stationary partial differential equations involving the spectral fractional Laplacian, and chooses a global Gaussian quadrature rule in discretization of the integral.
...
...

References

SHOWING 1-8 OF 8 REFERENCES
Accurate and efficient evaluation of the a posteriori error estimator in the reduced basis method
TLDR
A new approximation of the error bound using the Empirical Interpolation Method (EIM), which achieves higher levels of accuracy and requires potentially less precomputations than the usual formula.
REDUCED BASIS METHOD FOR FINITE VOLUME APPROXIMATIONS OF PARAMETRIZED LINEAR EVOLUTION EQUATIONS
TLDR
This work introduces a new offline basis-generation algorithm based on the derivation of rigorous a-posteriori error estimates in various norms for general linear evolution schemes such as finite volume schemes for parabolic and hyperbolic evolution equations.
Reduced Basis Approximation and A Posteriori Error Estimation for Parametrized Partial Differential Equations
The text is focused on formulation, analysis, and computational procedures for reduced basis approximation and a posteriori error estimation for parametrized PDEs in the real-time and many-query
Convergence Rates for Greedy Algorithms in Reduced Basis Methods
TLDR
The reduced basis method was introduced for the accurate online evaluation of solutions to a parameter dependent family of elliptic PDEs by determining a “good” n-dimensional space to be used in approximating the elements of a compact set $\mathcal{F}$ in a Hilbert space $\ mathscal{H}$.
A Space-Time Petrov-Galerkin Certified Reduced Basis Method: Application to the Boussinesq Equations
  • Masayuki Yano
  • Computer Science, Mathematics
    SIAM J. Sci. Comput.
  • 2014
We present a space-time certified reduced basis method for long-time integration of parametrized parabolic equations with quadratic nonlinearity which admit an affine decomposition in parameter but
The Reduced Basis Method for Time‐Harmonic Maxwell's Equations
TLDR
The Reduced Basis Method generates low‐order models of parametrized PDEs to allow for efficient evaluation of the input‐output behaviour in many‐query and real‐time contexts and plays a crucial role in the greedy process to generate the surrogate model.
Accurate a posteriori error evaluation in the reduced basis method
Reduced basis method for finite volume approximations of parametrized linear evolution equations . M 2 AN ( Math
  • . Model . Numer . Anal . )
  • 2008