H₂-optimal model reduction using projected nonlinear least squares
From MaRDI portal
Publication:5856679
Abstract: In many applications throughout science and engineering, model reduction plays an important role replacing expensive large-scale linear dynamical systems by inexpensive reduced order models that capture key features of the original, full order model. One approach to model reduction finds reduced order models that are locally optimal approximations in the norm, an approach taken by the Iterative Rational Krylov Algorithm (IRKA) among others. Here we introduce a new approach for -optimal model reduction using the projected nonlinear least squares framework previously introduced in [J. M. Hokanson, SIAM J. Sci. Comput. 39 (2017), pp. A3107--A3128]. At each iteration, we project the optimization problem onto a finite-dimensional subspace yielding a weighted least squares rational approximation problem. Subsequent iterations append this subspace such that the least squares rational approximant asymptotically satisfies the first order necessary conditions of the original, optimization problem. This enables us to build reduced order models with similar error in the norm but using far fewer evaluations of the expensive, full order model compared to competing methods. Moreover, our new algorithm only requires access to the transfer function of the full order model, unlike IRKA which requires a state-space representation or TF-IRKA which requires both the transfer function and its derivative. Applying the projected nonlinear least squares framework to the -optimal model reduction problem open new avenues for related model reduction problems.
Recommendations
- $\mathcal{H}_2$ Model Reduction for Large-Scale Linear Dynamical Systems
- A new framework for \(\mathcal{H}_2\)-optimal model reduction
- A projection-based algorithm for model-order reduction with \(H_2\) performance: a convex-optimization setting
- Frequency-weighted \(\mathcal{H}_2\)-optimal model order reduction via oblique projection
- Model reduction of large-scale systems by least squares
Cites work
- $\mathcal{H}_2$ Model Reduction for Large-Scale Linear Dynamical Systems
- A Subspace, Interior, and Conjugate Gradient Method for Large-Scale Bound-Constrained Minimization Problems
- A two-step model reduction approach for mechanical systems with moving loads
- ARPACK Users' Guide
- Accuracy and Stability of Numerical Algorithms
- Accurate Singular Value Decompositions of Structured Matrices
- Advanced Structural Dynamics and Active Control of Structures
- Approximation of Large-Scale Dynamical Systems
- Computing reduced order models via inner-outer Krylov recycling in diffuse optical tomography
- Convergence of the iterative rational Krylov algorithm
- Exponentially convergent Fourier-Chebyshev quadrature schemes on bounded and infinite intervals
- Inexact solves in interpolatory model reduction
- Interpolatory projection methods for structure-preserving model reduction
- Linear system approximation via covariance equivalent realizations
- Model Reduction for Large-Scale Systems with High-Dimensional Parametric Input Space
- Model reduction and approximation. Theory and algorithms
- Nonlinear parametric inversion using interpolatory model reduction
- Numerical Methods for Computing Angles Between Linear Subspaces
- Perturbation theory for linear operators.
- Pivoting and backward stability of fast algorithms for solving Cauchy linear equations
- Projected nonlinear least squares for exponential fitting
- Quadrature-based vector fitting for discretized \(\mathcal{H}_2\) approximation
- Rational interpolation and state-variable realizations
- Recycling BICG with an application to model reduction
- The AAA algorithm for rational approximation
- The Differentiation of Pseudo-Inverses and Nonlinear Least Squares Problems Whose Variables Separate
- Theory of Reproducing Kernels
- H₂-optimal model reduction of MIMO systems
- \({\mathcal H}_2\)-optimal model reduction with higher-order poles
Cited in
(11)- Preconditioned multishift BiCG for \(\mathcal{H}_2\)-optimal model reduction
- Preconditioning parametrized linear systems
- Two-Sided Projection Methods for Nonlinear Model Order Reduction
- \(\boldsymbol{\mathcal{L}_2}\)-Optimal Reduced-Order Modeling Using Parameter-Separable Forms
- A projection-based algorithm for model-order reduction with \(H_2\) performance: a convex-optimization setting
- Structure-Preserving $H^2$ Optimal Model Reduction Based on the Riemannian Trust-Region Method
- Homotopy methods for solving the optimal projection equations for theH2reduced order model problem
- Fast optimal \(\mathcal H_2\) model reduction algorithms based on Grassmann manifold optimization
- A new framework for \(\mathcal{H}_2\)-optimal model reduction
- A Unifying Framework for Interpolatory \({\boldsymbol{\mathcal{L}_2}}\)-Optimal Reduced-Order Modeling
- Galerkin v. least-squares Petrov-Galerkin projection in nonlinear model reduction
This page was built for publication: \( \mathcal{H}_2\)-optimal model reduction using projected nonlinear least squares
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5856679)
