The SPAI algorithm, a sparse approximate inverse preconditioning techniquefor large sparse linear systems, proposed by Grote and Huckle [SIAM J. Sci.Comput., 18 (1997), pp.~838--853.], is based on the F-norm minimization andcomputes a sparse approximate inverse $M$ of a large sparse matrix $A$adaptively. However, SPAI may be costly to seek the most profitable indices ateach loop and $M$ may be ineffective for preconditioning. In this paper, wepropose a residual based sparse approximate inverse preconditioning procedure(RSAI), which, unlike SPAI, is based on only the {\em dominant} rather than allinformation on the current residual and augments sparsity patterns adaptivelyduring the loops. RSAI is less costly to seek indices and is more effective tocapture a good approximate sparsity pattern of $A^{-1}$ than SPAI. To controlthe sparsity of $M$ and reduce computational cost, we develop a practicalRSAI($tol$) algorithm that drops small nonzero entries adaptively during theprocess. Numerical experiments are reported to demonstrate that RSAI($tol$) isat least competitive with SPAI and can be considerably more efficient andeffective than SPAI. They also indicate that RSAI($tol$) is comparable to thePSAI($tol$) algorithm proposed by one of the authors in 2009.
展开▼
机译:Grote and Huckle [SIAM J. Sci.Comput。,18(1997),pp。〜838--853。]提出的SPAI算法是一种用于大型稀疏线性系统的稀疏近似逆预处理技术。规范最小化并自适应地计算大稀疏矩阵$ A $的稀疏近似逆$ M $。但是,SPAI在每个循环中寻求最有利的指数可能会付出高昂的代价,而$ M $可能对预处理无效。在本文中,我们提出了一种基于残差的稀疏近似逆预处理程序(RSAI),与SPAI不同,该程序仅基于{\ em优势}而不是基于当前残差的所有信息,并在循环期间自适应地增强稀疏性模式。与SPAI相比,RSAI寻求索引的成本更低,并且能够更好地捕获$ A ^ {-1} $的良好近似稀疏模式。为了控制$ M $的稀疏性并减少计算成本,我们开发了一种实用的RSAI($ tol $)算法,该算法在处理过程中会自适应地丢弃小的非零条目。据报道,数值实验表明RSAI(toll $)至少可以与SPAI竞争,并且可以比SPAI更加有效。他们还表明RSAI($ tol $)与一位作者在2009年提出的PSAI($ tol $)算法相当。
展开▼