A Jacobi-like method for solving algebraic Riccati equations on parallel computers. (English) Zbl 0891.93033
A Jacobi-like method for computing the Hamiltonian-Schur form of a Hamiltonian matrix is suggested. The method is analogous to the Jacobi-like method of Eberlein for the computation of the Schur form of a general matrix. When used for Hermitian Hamiltonian matrices, the method is equivalent to the Kogbetliantz algorithm. Numerical experiments suggest a fast convergence. The method is used to solve continuous time Riccati equations.
Reviewer: S.Zlobec (Montreal)
MSC:
93B40 | Computational methods in systems theory (MSC2010) |
93B60 | Eigenvalue problems |
65F15 | Numerical computation of eigenvalues and eigenvectors of matrices |
15A24 | Matrix equations and identities |
65Y05 | Parallel numerical computation |