Generalization of the concept of diagonal dominance with applications to matrix D-stability
Introduction
Here, we introduce the following notations:
for the set of real matrices;
for the set of complex matrices;
for the spectrum of a matrix A (i.e. the set of all eigenvalues of A defined as zeroes of its characteristic polynomial );
for the open left-hand side of the complex plane , i.e.
for the closed right-hand side of the complex plane , i.e.
for the set of all positive diagonal matrices (i.e. matrices with positive entries on the principal diagonal while the entries outside the principal diagonal are zeroes);
for the subclass of , defined as follows:
for the subclass of , defined as follows:
In this paper, we generalize the concept of diagonal dominance. First, let us recall the classical definition of diagonally dominant matrices (see, for example, [7], also [15], [16]).
Definition 1 A matrix is called strictly row diagonally dominant if the following inequalities hold: A matrix A is called strictly column diagonally dominant if is strictly row diagonally dominant.
Definition 1'. A matrix is called generalized diagonally dominant if there exist positive scalars (weights) , such that (i.e. if there is a positive diagonal matrix , such that is strictly row diagonally dominant). If, in addition, , , then A is called negative diagonally dominant (NDD).
It is well-known (see, for example, [18], p. 382, Theorem 4.C.2) that if a complex matrix A is NDD, then A is Hurwitz stable, i.e. every satisfies . The stability of A implies the Lyapunov asymptotic stability of a system of ODE with the system matrix A. Being a sufficient for stability condition, negative diagonal dominance is of particular interest by itself due to its connection to the properties of nonlinear systems (see [17]). Less known is the fact that negative diagonal dominance implies some special concepts of matrix and system stability, that are stronger than just stability and shows stability preservation under specific matrix (system) perturbations. Such concepts include multiplicative D-stability (see, for example, [8], [10]).
Definition 2 A matrix A is called (multiplicative) D-stable if for all , where D is any matrix from .
Matrix D-stability plays an important role in the theory of stability (see, for example, [14], [9], [12] and references therein) and has numerous applications in the economics, mathematical ecology, mechanics and other branches of science. The proof of the fact that NDD matrices are D-stable is based on the simple observance that the negative diagonal dominance implies stability and is preserved under multiplication by a positive diagonal matrix (see [8], p. 54, Observation (i)). Note, that the property of diagonal dominance can be proved (disproved) by a finite number of steps while the conditions of D-stability of A involve checking all the products of the form , where D runs along the infinite set of positive diagonal matrices.
In this paper, we are concerned with the problem of a matrix spectra localization inside a prescribed convex region (so called -stability problem) and its robust aspects. Many problems of the system dynamics lead to establishing -stability of the system matrix with respect to a specified region (see, for example, [5], [6]). The regions of particular interest are LMI (Linear Matrix Inequality) regions, introduced in [3], that include the shifted left half-plane, the unit disk and the conic sector around the negative direction of the real axis. Due to the rapid advance in control theory and its application, we face the problem of finding some easy-to-verify conditions, which allow us to establish:
- -
-stability of a given matrix A;
- -
the preservation of -stability under some specific perturbations of A, e.g. multiplication by a diagonal matrix.
Though an amount of research (see [5], [6]) is due to -stability (or eigenvalue clustering in a region ), the progress in describing the matrix classes which are -stable, and, moreover, preserve -stability under multiplication by a specific diagonal matrix is still very little and concerns classical regions such as the left half-plane (see [1], [4], [8]) and the unit disk (see [9]). Here, we make certain efforts to fill this gap.
The paper is organized as follows. In Section 2, we study unbounded LMI regions and their properties, focusing on the most well-known examples, such as the shifted left half-plane, the conic sector, hyperbola and parabola. We also recall the concept of -stability for an unbounded LMI region . In Section 3, we introduce the crucial concept of this paper, namely, diagonal dominance with respect to a given LMI region (so-called diagonal -dominance). We consider the particular cases of diagonal -dominance, with respect to the most important LMI regions. Section 4 deals with the basic properties of diagonally -dominant matrices. Section 5 gives the most important results of the paper, i.e. the implications between diagonal -dominance, -stability and -stability.
Section snippets
Definition and examples of LMI regions
Consider the following type of regions, introduced in [2] (see also [3]).
Definition 3 A subset that can be defined as where , , is called an LMI region with the characteristic function and generating matrices M and L. It was shown in [11], that an LMI region is open. In the sequel, we shall use the notation for its closure and for its boundary. Later on, we shall be especially interested in the following types of LMI regions.
Example 1 The shifted left
General definition
Given a (bounded or unbounded) LMI region , let the intersection of with the real axis be denoted by where both α and β may be infinite. According to this notation, . Note, that is nonempty whenever is nonempty (see [11], Lemma 21).
Let the part of the boundary of which lies above the real axis (if is bounded from above) be represented by a concave function . Note, that the boundary function of an LMI region , is defined implicitly.
We define
Properties of diagonally -dominant matrices
Here, we consider the basic properties of diagonally -dominant matrices, that we shall use later. First, let us state and prove the following technical lemma. Lemma 1 Let the function be defined by the formula: Then whenever , . Proof Since and , we obtain the following inequalities: The last
Diagonal -dominance implies -stability and -stability
First, we recall the following fundamental statement of matrix eigenvalue localization (see, for example, [7], p. 344). Theorem 1 Gershgorin Let , define Let be a closed disk centered at with the radius . Then all the eigenvalues of A are located in the union of n discs
Let us recall a well-known fact that if a complex matrix is
Conclusions
In this paper, we generalized the concept of diagonally dominance, which is known to be a sufficient condition for stability of a given matrix A. In particular, we addressed the problem of a matrix spectra localization inside a prescribed convex region (the so called -stability problem), with the purpose of finding some easy-to-verify conditions, so as to establish:
- (i)
-stability of a given matrix A;
- (ii)
the preservation of -stability under some specific perturbations of A, e.g. multiplication by
Declaration of Competing Interest
No competing interest.
Acknowledgements
The research was supported by the National Science Foundation of China grant number 12050410229.
References (18)
Three types of matrix stability
Linear Algebra Appl.
(1978)Sufficient conditions for D-stability
Int. J. Econ. Theory
(1974)Stronger-than-Lyapunov notions of matrix stability, or how “flowers” help solve problems in mathematical ecology
Linear Algebra Appl.
(2005)- et al.
Stability of matrices with negative diagonal submatrices
Linear Algebra Appl.
(2002) - et al.
Stability of matrices with sufficiently strong negative-dominant-diagonal submatrices
Linear Algebra Appl.
(1997) - et al.
A note on dynamical stability
Econometrica
(1958) - et al.
design with pole placement constraints: an LMI approach
IEEE Trans. Autom. Control
(1996) - et al.
Robust pole placement in LMI regions
Root Clustering in Parameter Space
(1990)
Cited by (4)
Some bounds for determinants of relatively D-stable matrices
2023, Linear Algebra and Its ApplicationsCitation Excerpt :Under the assumptions of Theorem 10, more precise spectra localization can be given. Let us recall the following definition (see [12], Definition 6'). Let us also recall a well-known definition of diagonally dominant matrices.
Novel Versions of D-Stability in Matrices Provide New Insights into ODE Dynamics
2023, Mediterranean Journal of MathematicsStability and minimal decay rate of systems of ODE
2022, Research Square