Optimal reduced model algorithms for data-based state estimation

DOI10.1137/19M1255185MaRDI QIDQ3386988zbMATH OpenOpenAlexFDO

Authors

Publication date 12 January 2021

Published in SIAM Journal on Numerical Analysis (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1903.07938

zbMATH Keywords

convex optimizationoptimal recoveryparametrized PDEsreduced modelssensing

Mathematics Subject Classification ID

Inverse problems for PDEs (35R30) Numerical approximation and computational geometry (primarily algorithms) (65D99) Numerical optimization and variational techniques (65K10) Error bounds for initial value and initial-boundary value problems involving PDEs (65M15) Numerical methods for inverse problems for initial value and initial-boundary value problems involving PDEs (65M32)

Abstract: Reduced model spaces, such as reduced basis and polynomial chaos, are linear spaces $V_n$ of finite dimension $n$ which are designed for the efficient approximation of families parametrized PDEs in a Hilbert space $V$. The manifold $mathcalM$ that gathers the solutions of the PDE for all admissible parameter values is globally approximated by the space $V_n$ with some controlled accuracy $epsilo{n}_n$, which is typically much smaller than when using standard approximation spaces of the same dimension such as finite elements. Reduced model spaces have also been proposed in [13] as a vehicle to design a simple linear recovery algorithm of the state $uinmathcalM$ corresponding to a particular solution when the values of parameters are unknown but a set of data is given by $m$ linear measurements of the state. The measurements are of the form $el{l}_j(u)$, $j=1,dots,m$, where the $el{l}_j$ are linear functionals on $V$. The analysis of this approach in [2] shows that the recovery error is bounded by $m{u}_{n}epsilo{n}_n$, where $m{u}_n=mu(V_n,W)$ is the inverse of an inf-sup constant that describe the angle between $V_n$ and the space $W$ spanned by the Riesz representers of $(el{l}_1,dots,el{l}_m)$. A reduced model space which is efficient for approximation might thus be ineffective for recovery if $m{u}_n$ is large or infinite. In this paper, we discuss the existence and construction of an optimal reduced model space for this recovery method, and we extend our search to affine spaces. Our basic observation is that this problem is equivalent to the search of an optimal affine algorithm for the recovery of $mathcalM$ in the worst case error sense. This allows us to perform our search by a convex optimization procedure. Numerical tests illustrate that the reduced model spaces constructed from our approach perform better than the classical reduced basis spaces.

Recommendations

• Nonlinear reduced models for state and parameter estimation
• Greedy Algorithms for Optimal Measurements Selection in State Estimation Using Reduced Models
• Nonlinear reduced DNN models for state estimation
• Nonlinear methods for model reduction
• Adaptive reduced basis methods for parameterized evolution problems with application in optimization and state estimation

Cites work

• \textit{A priori} convergence of the greedy algorithm for the parametrized reduced basis method
• A first-order primal-dual algorithm for convex problems with applications to imaging
• A parameterized-background data-weak approach to variational data assimilation: formulation, analysis, and application to acoustics
• An adaptive parametrized-background data-weak approach to variational data assimilation
• Analytic regularity and polynomial approximation of parametric and stochastic elliptic PDE's
• Approximation of high-dimensional parametric PDEs
• Convergence rates for greedy algorithms in reduced basis methods
• Data Assimilation
• Data Assimilation in Reduced Modeling
• Greedy Algorithms for Optimal Measurements Selection in State Estimation Using Reduced Models
• Greedy algorithms for reduced bases in Banach spaces
• scientific article; zbMATH DE number 4144656 (Why is no real title available?)
• scientific article; zbMATH DE number 727440 (Why is no real title available?)
• Inverse problems: a Bayesian perspective
• Nonlinear methods for model reduction
• Nonlinear model order reduction based on local reduced-order bases
• Reduced basis approximation and a posteriori error bounds for 4D-Var data assimilation
• Reduced basis approximation and a posteriori error estimation for affinely parametrized elliptic coercive partial differential equations. Application to transport and continuum mechanics.
• Reduced Basis Greedy Selection Using Random Training Sets
• Survey of multifidelity methods in uncertainty propagation, inference, and optimization
• Tractability of multivariate problems. Volume I: Linear information

Cited in

(24)

• Data assimilation and sampling in Banach spaces
• Data-driven robust state estimation for reduced-order models of 2D Boussinesq equations with parametric uncertainties
• A real-time variational data assimilation method with data-driven model enrichment for time-dependent problems
• Greedy Algorithms for Optimal Measurements Selection in State Estimation Using Reduced Models
• Accuracy controlled data assimilation for parabolic problems
• State estimation with model reduction and shape variability. Application to biomedical problems
• Nonlinear reduced DNN models for state estimation
• Instances of computational optimal recovery: dealing with observation errors
• Data Assimilation in Reduced Modeling
• Nonlinear reduced models for state and parameter estimation
• Nonlinear approximation spaces for inverse problems
• Impact of physical model error on state estimation for neutronics applications
• A greedy sensor selection algorithm for hyperparameterized linear Bayesian inverse problems with correlated noise models
• State Estimation—The Role of Reduced Models
• Dictionary-based model reduction for state estimation
• Model reduction from partial observations
• Inverse problems: a deterministic approach using physics-based reduced models
• Neural-network-augmented empirical interpolation method for field reconstruction with noise and vibration tolerance
• Identification of active component functions in finite-max minimisation via a smooth reformulation
• Introduction: Wolfgang Dahmen's mathematical work (as of 2009)
• Dynamical approximation and sensor placement for filtering problems
• On the criteria for online updating of basis functions for model order reduction techniques
• Some new convergence analysis and applications of POD-Greedy algorithms
• Symplectic filtering of Hamiltonian dynamics with moving sensors

This page was built for publication: Optimal reduced model algorithms for data-based state estimation

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3386988)