Factorized solution of Lyapunov equations based on hierarchical matrix arithmetic (Q858172): Difference between revisions
From MaRDI portal
Changed an Item |
Changed an Item |
||
Property / describes a project that uses | |||
Property / describes a project that uses: hlib / rank | |||
Normal rank |
Revision as of 11:35, 28 February 2024
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Factorized solution of Lyapunov equations based on hierarchical matrix arithmetic |
scientific article |
Statements
Factorized solution of Lyapunov equations based on hierarchical matrix arithmetic (English)
0 references
8 January 2007
0 references
To obtain the symmetric positive definite solution \(X\) of a Lyapunov equation \(AX+XA^T+BB^T=0\), it is best to compute its factor \(Y\) such that \(X=YY^T\). An approximation to \(Y\) can be obtained using the sign method [\textit{J. D. Roberts}, Int. J. Control 32, 677--687 (1980, Zbl 0463.93050)]. In some applications (finite element method; boundary element method) the matrices have a hierarchical structure of blocks of low rank (so called \(\mathcal H\) matrices). In this paper a method is proposed that combines the \(\mathcal H\) matrix structure with the sign method to compute a \(Y\)-factor of the solution of the Lyapunov equation. An extension to more general Lyapunov equations (such as those appearing in descriptor systems) is also included.
0 references
H-matrix
0 references
hierarchical matrix
0 references
rank structured matrix
0 references
sign function
0 references
model reduction
0 references
Lyapunov equation
0 references
descriptor system
0 references
sign method
0 references