Krylov methods for low‐rank commuting generalized Sylvester equations

Abstract

Summary We consider generalizations of the Sylvester matrix equation, consisting of the sum of a Sylvester operator and a linear operator Π with a particular structure. More precisely, the commutators of the matrix coefficients of the operator Π and the Sylvester operator coefficients are assumed to be matrices with low rank. We show (under certain additional conditions) low‐rank approximability of this problem, that is, the solution to this matrix equation can be approximated with a low‐rank matrix. Projection methods have successfully been used to solve other matrix equations with low‐rank approximability. We propose a new projection method for this class of matrix equations. The choice of the subspace is a crucial ingredient for any projection method for matrix equations. Our method is based on an adaption and extension of the extended Krylov subspace method for Sylvester equations. A constructive choice of the starting vector/block is derived from the low‐rank commutators. We illustrate the effectiveness of our method by solving large‐scale matrix equations arising from applications in control theory and the discretization of PDEs. The advantages of our approach in comparison to other methods are also illustrated.

New Gramians for Switched Linear Systems: Reachability, Observability, and Model Reduction

“…However, in this Kronecker form, the solution of the generalized Lyapunov equation is determined by solving a set of n(n + 1)/2 equations in n(n + 1)/2 variables, whose cost is O(n 6 ) operations. Fortunately, new efficient methodologies have been developed recently to determine low-rank solutions of these generalized Lyapunov equations (see [19], [12], [34] and [25]) which are suitable in the large-scale setting.…”

Residual-based iterations for the generalized Lyapunov equation

“…All algorithms are treated in a subspace fashion and we compare practically achieved approximation properties as a function of subspace dimension. For small and moderate sized problems there are algorithms for computing the full solution, although costly, see [23,Algorithm 2], cf. [33, equation (12)].…”

Optimality Properties of Galerkin and Petrov–Galerkin Methods for Linear Matrix Equations

“…Sometimes this is the only option as effective algorithms to solve (1.2) in its natural matrix equation form are still lacking in the literature in the most general case. The methods developed so far require some additional assumptions on the coefficient matrices A j , B j ; see, e.g., [7,17,20,30,34]. In this section we show that exploiting the matrix structure of equation (1.2) not only leads to numerical algorithms with lower computational costs per iteration and modest storage demands, but they also avoid some spectral redundancy encoded in the problem formulation (4.1).…”

“…Without further hypotheses, the symmetric matrix Y k solving (5.5) is indefinite in general, thus preventing Y k from preserving the semidefiniteness property of the solution to be approximated. (N(·)) is less than one; see, e.g., [11,17,34] for various implementations.…”