Variational Analysis of the Spectral Abscissa at a Matrix with a Nongeneric Multiple Eigenvalue

Sara Grundel, Michael L. Overton

Research output: Contribution to journalArticle

Abstract

The spectral abscissa is a fundamental map from the set of complex matrices to the real numbers. Denoted α and defined as the maximum of the real parts of the eigenvalues of a matrix X, it has many applications in stability analysis of dynamical systems. The function α is nonconvex and is non-Lipschitz near matrices with multiple eigenvalues. Variational analysis of this function was presented in Burke and Overton (Math Program 90:317-352, 2001), including a complete characterization of its regular subgradients and necessary conditions which must be satisfied by all its subgradients. A complete characterization of all subgradients of α at a matrix X was also given for the case that all active eigenvalues of X (those whose real part equals α(X)) are nonderogatory (their geometric multiplicity is one) and also for the case that they are all nondefective (their geometric multiplicity equals their algebraic multiplicity). However, necessary and sufficient conditions for all subgradients in all cases remain unknown. In this paper we present necessary and sufficient conditions for the simplest example of a matrix X with a derogatory, defective multiple eigenvalue.

Original languageEnglish (US)
Pages (from-to)19-43
Number of pages25
JournalSet-Valued and Variational Analysis
Volume22
Issue number1
DOIs
StatePublished - Mar 2014

Fingerprint

Abscissa
Multiple Eigenvalues
Variational Analysis
Subgradient
Multiplicity
Necessary Conditions
Eigenvalue
Non-Lipschitz
Sufficient Conditions
Stability Analysis
Dynamical systems
Dynamical system
Unknown

Keywords

  • Degenerate eigenvalue
  • Derogatory eigenvalue
  • Spectral abscissa
  • Subgradient
  • Variational analysis

ASJC Scopus subject areas

  • Analysis
  • Applied Mathematics
  • Geometry and Topology
  • Numerical Analysis
  • Statistics and Probability

Cite this

Variational Analysis of the Spectral Abscissa at a Matrix with a Nongeneric Multiple Eigenvalue. / Grundel, Sara; Overton, Michael L.

In: Set-Valued and Variational Analysis, Vol. 22, No. 1, 03.2014, p. 19-43.

Research output: Contribution to journalArticle

@article{ce1c3287c0734f2ca2d357f74da905ee,
title = "Variational Analysis of the Spectral Abscissa at a Matrix with a Nongeneric Multiple Eigenvalue",
abstract = "The spectral abscissa is a fundamental map from the set of complex matrices to the real numbers. Denoted α and defined as the maximum of the real parts of the eigenvalues of a matrix X, it has many applications in stability analysis of dynamical systems. The function α is nonconvex and is non-Lipschitz near matrices with multiple eigenvalues. Variational analysis of this function was presented in Burke and Overton (Math Program 90:317-352, 2001), including a complete characterization of its regular subgradients and necessary conditions which must be satisfied by all its subgradients. A complete characterization of all subgradients of α at a matrix X was also given for the case that all active eigenvalues of X (those whose real part equals α(X)) are nonderogatory (their geometric multiplicity is one) and also for the case that they are all nondefective (their geometric multiplicity equals their algebraic multiplicity). However, necessary and sufficient conditions for all subgradients in all cases remain unknown. In this paper we present necessary and sufficient conditions for the simplest example of a matrix X with a derogatory, defective multiple eigenvalue.",
keywords = "Degenerate eigenvalue, Derogatory eigenvalue, Spectral abscissa, Subgradient, Variational analysis",
author = "Sara Grundel and Overton, {Michael L.}",
year = "2014",
month = "3",
doi = "10.1007/s11228-013-0234-7",
language = "English (US)",
volume = "22",
pages = "19--43",
journal = "Set-Valued and Variational Analysis",
issn = "1877-0533",
publisher = "Springer Verlag",
number = "1",

}

TY - JOUR

T1 - Variational Analysis of the Spectral Abscissa at a Matrix with a Nongeneric Multiple Eigenvalue

AU - Grundel, Sara

AU - Overton, Michael L.

PY - 2014/3

Y1 - 2014/3

N2 - The spectral abscissa is a fundamental map from the set of complex matrices to the real numbers. Denoted α and defined as the maximum of the real parts of the eigenvalues of a matrix X, it has many applications in stability analysis of dynamical systems. The function α is nonconvex and is non-Lipschitz near matrices with multiple eigenvalues. Variational analysis of this function was presented in Burke and Overton (Math Program 90:317-352, 2001), including a complete characterization of its regular subgradients and necessary conditions which must be satisfied by all its subgradients. A complete characterization of all subgradients of α at a matrix X was also given for the case that all active eigenvalues of X (those whose real part equals α(X)) are nonderogatory (their geometric multiplicity is one) and also for the case that they are all nondefective (their geometric multiplicity equals their algebraic multiplicity). However, necessary and sufficient conditions for all subgradients in all cases remain unknown. In this paper we present necessary and sufficient conditions for the simplest example of a matrix X with a derogatory, defective multiple eigenvalue.

AB - The spectral abscissa is a fundamental map from the set of complex matrices to the real numbers. Denoted α and defined as the maximum of the real parts of the eigenvalues of a matrix X, it has many applications in stability analysis of dynamical systems. The function α is nonconvex and is non-Lipschitz near matrices with multiple eigenvalues. Variational analysis of this function was presented in Burke and Overton (Math Program 90:317-352, 2001), including a complete characterization of its regular subgradients and necessary conditions which must be satisfied by all its subgradients. A complete characterization of all subgradients of α at a matrix X was also given for the case that all active eigenvalues of X (those whose real part equals α(X)) are nonderogatory (their geometric multiplicity is one) and also for the case that they are all nondefective (their geometric multiplicity equals their algebraic multiplicity). However, necessary and sufficient conditions for all subgradients in all cases remain unknown. In this paper we present necessary and sufficient conditions for the simplest example of a matrix X with a derogatory, defective multiple eigenvalue.

KW - Degenerate eigenvalue

KW - Derogatory eigenvalue

KW - Spectral abscissa

KW - Subgradient

KW - Variational analysis

UR - http://www.scopus.com/inward/record.url?scp=84893593261&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84893593261&partnerID=8YFLogxK

U2 - 10.1007/s11228-013-0234-7

DO - 10.1007/s11228-013-0234-7

M3 - Article

VL - 22

SP - 19

EP - 43

JO - Set-Valued and Variational Analysis

JF - Set-Valued and Variational Analysis

SN - 1877-0533

IS - 1

ER -