On the Positive Semidefinite and Skew-Hermitian Splitting Preconditioner for Generalized Saddle Point Problems
We consider the positive semidefinite and skew-Hermitian splitting (PSS) preconditioner for generalized saddle point problems with nonzero (2,2) blocks. It is shown that all eigenvalues of the PSS preconditioned matrix form two tight clusters, one is near (0,0) and the other is near (2,0) when the iteration parameter approaches to zero from above. This property explains why the optimal iteration parameter leading to the fastest convergence rate is usually small. The model problem of Oseen equations is used to illustrate the presented spectral properties as well as the performance of the PSS preconditioner.
Saddle point problem preconditioner eigenvalue
Shu-Qian Shen Guang-Bin Wang
School of Mathematics and Computational Sciences, China University of Petroleum, Dongying Shandong, Department of Mathematics, Qingdao University of Science and Technology, Qingdao Shandong, 266061, P
国际会议
The Third International Workshop on Applied Matriz Theory(第三届国际矩阵分析与应用会议)
杭州
英文
941-944
2009-07-09（万方平台首次上网日期，不代表论文的发表时间）