Mixed forward-backward stability of the two-level orthogonal Arnoldi method for quadratic problems

DOI10.1016/J.LAA.2018.04.020zbMATH Open1391.65087arXiv1707.00930OpenAlexW2963815419MaRDI QIDQ1641996FDOQ1641996

Authors: Karl Meerbergen, Javier Pérez 

Publication date: 20 June 2018

Published in: Linear Algebra and its Applications (Search for Journal in Brave)

Abstract: We revisit the numerical stability of the two-level orthogonal Arnoldi (TOAR) method for computing an orthonormal basis of a second--order Krylov subspace associated with two given matrices. We show that the computed basis is close (on certain subspace metric sense) to a basis for a second-order Krylov subspace associated with nearby coefficient matrices, provided that the norms of the given matrices are not too large or too small. Thus, the results in this work provide for the first time conditions that guarantee the numerical stability of the TOAR method in computing orthonormal bases of second-order Krylov subspaces. We also study scaling the quadratic problem for improving the numerical stability of the TOAR procedure when the norms of the matrices are too large or too small. We show that in many cases the TOAR procedure applied to scaled matrices is numerically stable when the scaling introduced by Fan, Lin and Van Dooren is used.

Full work available at URL: https://arxiv.org/abs/1707.00930

Recommendations

• Stability analysis of the two-level orthogonal Arnoldi procedure
• SOAR: A Second-order Arnoldi Method for the Solution of the Quadratic Eigenvalue Problem
• Implicitly restarted two-level orthogonal Arnoldi algorithms
• The Quadratic Arnoldi Method for the Solution of the Quadratic Eigenvalue Problem
• A variant of second-order Arnoldi method for solving the quadratic eigenvalue problem

zbMATH Keywords

numerical stabilityKrylov subspaceArnoldi algorithmsecond-order Arnoldi algorithmsecond-order Krylov subspacetwo-level orthogonal Arnoldi algorithm

Mathematics Subject Classification ID

Numerical computation of eigenvalues and eigenvectors of matrices (65F15)

Cites Work

• An algorithm for the complete solution of quadratic eigenvalue problems
• Reorthogonalization and Stable Algorithms for Updating the Gram-Schmidt QR Factorization
• Title not available (Why is that?)
• Multiplicative perturbation theory of the Moore-Penrose inverse and the least squares problem
• Title not available (Why is that?)
• Dimension Reduction of Large-Scale Second-Order Dynamical Systems via a Second-Order Arnoldi Method
• A memory-efficient model order reduction for time-delay systems
• SOAR: A Second-order Arnoldi Method for the Solution of the Quadratic Eigenvalue Problem
• The eigenstructure of an arbitrary polynomial matrix: Computational aspects
• Normwise Scaling of Second Order Polynomial Matrices
• The Quadratic Arnoldi Method for the Solution of the Quadratic Eigenvalue Problem
• Parallel Krylov solvers for the polynomial eigenvalue problem in SLEPc
• Memory-efficient Arnoldi algorithms for linearizations of matrix polynomials in Chebyshev basis.
• Compact rational Krylov methods for nonlinear eigenvalue problems
• On the stability of computing polynomial roots via confederate linearizations
• Chebyshev rootfinding via computing eigenvalues of colleague matrices: when is it stable?
• Stability analysis of the two-level orthogonal Arnoldi procedure
• On the convergence of Ritz pairs and refined Ritz vectors for quadratic eigenvalue problems

Cited In (3)

• Implicitly restarted two-level orthogonal Arnoldi algorithms
• Krylov type methods for linear systems exploiting properties of the quadratic numerical range
• Stability analysis of the two-level orthogonal Arnoldi procedure

Uses Software

• mctoolbox
• quadeig

This page was built for publication: Mixed forward-backward stability of the two-level orthogonal Arnoldi method for quadratic problems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1641996)