Abstract:For nonsymmetric saddle point problems arising from the steady Navier-Stokes equations, Pan, Ng and Bai presented a deteriorated positive-definite and skew-Hermitian splitting (DPSS) preconditioner (Pan et al., 2006) to accelerate the convergence rates of the Krylov subspace iteration methods such as GMRES. In this paper, the unconditional convergence property of the DPSS iteration method is proved and a relaxed DPSS (RDPSS) preconditioner is proposed. The RDPSS preconditioner is much closer to the coefficient matrix than the DPSS preconditioner in certain norm, which straightforwardly results in an RDPSS iteration method. The convergence conditions of the RDPSS iteration are analyzed and the optimal parameter, which minimizes the spectral radius of the RDPSS iteration matrix, is derived. Using the RDPSS preconditioner to accelerate some Krylov subspace methods (like GMRES) is also studied. The eigenproperty of the preconditioned matrix is described and an upper bound of the degree of the minimal polynomial of the preconditioned matrix is obtained. Finally, numerical experiments of a model Navier-Stokes equation are presented to illustrate the efficiency of the RDPSS preconditioner. (C) 2014 Elsevier B.V. All rights reserved.

A Modified SSOR-like Preconditioner for Non-Hermitian Positive Definite Matrices

Preconditioning Technique for Symmetric M-matrices

Hermitian Preconditioning for a class of Non-Hermitian Linear Systems

A CONSTRAINT PRECONDITIONER FOR SOLVING SYMMETRIC POSITIVE DEFINITE SYSTEMS AND APPLICATION TO THE HELMHOLTZ EQUATIONS AND POISSON EQUATIONS

Shift-splitting preconditioner for non-Hermitian positive definite matrices

A Note on Preconditioning for M-Matrix

A Relaxed Deteriorated Pss Preconditioner for Nonsymmetric Saddle Point Problems from the Steady Navier-Stokes Equation

Preconditioner Design via Bregman Divergences

Regularized DPSS Preconditioners for Generalized Saddle Point Linear Systems

Computing the Hermitian Positive Definite Solutions of a Nonlinear Matrix Equation

A Generalization of T. Chan’s Preconditioner

Generalized Preconditioned Locally Harmonic Residual Method for Non-Hermitian Eigenproblems

A modified version of the PRESB preconditioner for a class of non-Hermitian complex systems of linear equations

A projected preconditioned conjugate gradient algorithm for computing many extreme eigenpairs of a Hermitian matrix

A Robust Two-Level Schwarz Preconditioner For Sparse Matrices

Preconditioner Design via the Bregman Divergence

Alternating positive semidefinite splitting preconditioners for double saddle point problems

Efficient Preconditioner Updates for Unsymmetric Shifted Linear Systems

Extended Newton-type Method for Inverse Singular Value Problems with Multiple And/or Zero Singular Values

Preconditioned AOR Iterative Method for H-matrices

On the Hermitian positive definite solution and Newton's method for a nonlinear matrix equation