Comparison results on preconditioned SOR-type iterative method for Z-matrices linear systems

From MaRDI portal
Publication:2370685





The aim of the material is to present some comparison theorems on the preconditioned successive overrelaxation (SOR)-type iterative method for \(Z\)-matrices linear systems. Convergence of the method applied to a \(Z\)-matrix system are discussed. The first section represents an introduction concerning the iterative methods for numerically solving linear systems \(Ax=b\) where \(A\) is an \(n\times n\) square matrix, and \(x\) and \(b\) are \(n\)-dimensional vectors. The classical SOR method, the Gauss-Seidel method, the preconditioned SOR-type iterative method [cf. \textit{T.-Z. Huang, G.-H. Cheng} and \textit{X.-Y. Cheng}, Appl. Math. Comput. 175, No. 1, 258--268 (2006; Zbl 1095.65024)] and the preconditioned Gauss-Seidel iterative method [cf. \textit{T. Z. Huang} and \textit{R. S. Ran}, J. Comput. Appl. Math. 177, No. 2, 455--459 (2005; Zbl 1073.15016)] for solving irreductible \(Z\)-matrices linear systems are presented. The second section contains the main comparison results with the SOR-type method, given by seven theorems and three corollaries. Thus, the authors conclude that the rate of convergence of the preconditioned Gauss-Seidel method is faster than the rate of convergence of the preconditioned SOR iterative method. A numerical experiment performed for randomly generated nonsingular \(M\)-matrices for \(n=10,\) shows that the results are true.











This page was built for publication: Comparison results on preconditioned SOR-type iterative method for \(Z\)-matrices linear systems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2370685)