Subspace Methods for Computing the Pseudospectral Abscissa and the Stability Radius

The pseudospectral abscissa and the stability radius are well-established tools for quantifying the stability of a matrix under unstructured perturbations. Based on first-order eigenvalue expansions, Guglielmi and Overton [SIAM J. Matrix Anal. Appl., 32 (2011), pp. 1166--1192] recently proposed a linearly converging iterative method for computing the pseudospectral abscissa. In this paper, we propose to combine this method and its variants with subspace acceleration. Each extraction step computes the pseudospectral abscissa of a small rectangular matrix pencil, which is comparably cheap and guarantees monotonicity. We observe local quadratic convergence and prove local superlinear convergence of the resulting subspace methods. Moreover, these methods extend naturally to computing the stability radius. A number of numerical experiments demonstrate the robustness and efficiency of the subspace methods.

[1]  Diederich Hinrichsen,et al.  Mathematical Systems Theory I , 2006, IEEE Transactions on Automatic Control.

[2]  Michael Elad,et al.  The Generalized Eigenvalue Problem for Nonsquare Pencils Using a Minimal Perturbation Approach , 2005, SIAM J. Matrix Anal. Appl..

[3]  Daniel Kressner.  Finding the distance to instability of a large sparse matrix , 2006, 2006 IEEE Conference on Computer Aided Control System Design, 2006 IEEE International Conference on Control Applications, 2006 IEEE International Symposium on Intelligent Control.

[4]  R. Byers.  A Bisection Method for Measuring the Distance of a Stable Matrix to the Unstable Matrices , 1988 .

[5]  L. Trefethen.  Spectra and pseudospectra , 2005 .

[6]  Nicola Guglielmi,et al.  Computing the Structured Pseudospectrum of a Toeplitz Matrix and Its Extreme Points , 2012, SIAM J. Matrix Anal. Appl..

[7]  Diederich Hinrichsen,et al.  An algorithm for the computation of the structured complex stability radius , 1989, Autom..

[8]  A. Bunse-Gerstner,et al.  Numerical computation of an analytic singular value decomposition of a matrix valued function , 1991 .

[9]  Nicola Guglielmi,et al.  Low-Rank Dynamics for Computing Extremal Points of Real Pseudospectra , 2013, SIAM J. Matrix Anal. Appl..

[10]  J. Walsh.  Interpolation and Approximation by Rational Functions in the Complex Domain , 1935 .

[11]  P. Wedin.  Perturbation theory for pseudo-inverses , 1973 .

[12]  C. D. Meyer,et al.  Derivatives and perturbations of eigenvectors , 1988 .

[13]  E. Gekeler,et al.  On the solution of systems of equations by the epsilon algorithm of Wynn , 1972 .

[14]  M. Steinbuch,et al.  A fast algorithm to computer the H ∞ -norm of a transfer function matrix , 1990 .

[15]  Joseph F. Grcar,et al.  Operator Coefficient Methods for Linear Equations , 2012, 1203.2390.

[16]  G. Stewart.  Matrix Algorithms, Volume II: Eigensystems , 2001 .

[17]  Adrian S. Lewis,et al.  Variational Analysis of Pseudospectra , 2008, SIAM J. Optim..

[18]  Sabine Van Huffel,et al.  SLICOT—A Subroutine Library in Systems and Control Theory , 1999 .

[19]  G. Alistair Watson,et al.  An Algorithm for Computing the Distance to Instability , 1998, SIAM J. Matrix Anal. Appl..

[20]  L. Trefethen,et al.  Spectra and pseudospectra : the behavior of nonnormal matrices and operators , 2005 .

[21]  M. Overton,et al.  FAST APPROXIMATION OF THE H∞ NORM VIA OPTIMIZATION OVER SPECTRAL VALUE SETS∗ , 2012 .

[22]  Michael L. Overton,et al.  Fast Algorithms for the Approximation of the Pseudospectral Abscissa and Pseudospectral Radius of a Matrix , 2011, SIAM J. Matrix Anal. Appl..

[23]  Chao Yang,et al.  ARPACK users' guide - solution of large-scale eigenvalue problems with implicitly restarted Arnoldi methods , 1998, Software, environments, tools.

[24]  D. Hinrichsen,et al.  Optimization problems in the robustness analysis of linear state space systems , 1989 .

[25]  M. Overton,et al.  Algorithms for the computation of the pseudospectral radius and the numerical radius of a matrix , 2005 .

[26]  G. W. Stewart,et al.  Matrix algorithms , 1998 .

[27]  A. Lewis,et al.  Robust stability and a criss‐cross algorithm for pseudospectra , 2003 .

[28]  L. Trefethen,et al.  Pseudospectra of rectangular matrices , 2002, Spectra and Pseudospectra.

[29]  Charles R. Johnson,et al.  Matrix analysis , 1985, Statistical Inference for Engineers and Data Scientists.

[30]  F. Potra.  OnQ-order andR-order of convergence , 1989 .

[31]  Jack Dongarra,et al.  Templates for the Solution of Algebraic Eigenvalue Problems , 2000, Software, environments, tools.

[32]  James M. Ortega,et al.  Iterative solution of nonlinear equations in several variables , 2014, Computer science and applied mathematics.

[33]  A. Sidi,et al.  Extrapolation methods for vector sequences , 1987 .

[34]  Michael L. Overton,et al.  Some Regularity Results for the Pseudospectral Abscissa and Pseudospectral Radius of a Matrix , 2012, SIAM J. Optim..

[35]  F. Leibfritz.  COMPleib: COnstrained Matrix–optimization Problem library – a collection of test examples for nonlinear semidefinite programs, control system design and related problems , 2006 .

[36]  Stephen J. Wright,et al.  Numerical Optimization , 2018, Fundamental Statistical Inference.

[37]  Alastair Spence,et al.  A Newton-based method for the calculation of the distance to instability , 2011 .

[38]  Stephen P. Boyd,et al.  A bisection method for computing the H∞ norm of a transfer matrix and related problems , 1989, Math. Control. Signals Syst..

[39]  Tosio Kato.  Perturbation theory for linear operators , 1966 .

[40]  S. Boyd,et al.  A regularity result for the singular values of a transfer matrix and a quadratically convergent algorithm for computing its L ∞ -norm , 1990 .

[41]  Michael L. Overton,et al.  Fast Approximation of the HINFINITY Norm via Optimization over Spectral Value Sets , 2013, SIAM J. Matrix Anal. Appl..

[42]  P. Wedin.  Perturbation bounds in connection with singular value decomposition , 1972 .

[43]  Jorge J. Moré,et al.  Digital Object Identifier (DOI) 10.1007/s101070100263 , 2001 .