对称正半定组的多项式加速方法  

POLYNOMIAL ACCELERATION METHODS FOR SYMMETRIC POSITIVE SEMIDEFINITE SYSTEMS

作　　者：曹志浩[1] 

机构地区：[1]复旦大学

出　　处：《高等学校计算数学学报》1990年第2期167-176,共10页Numerical Mathematics A Journal of Chinese Universities

基　　金：博士点基金

摘　　要：§1.引言 考虑(可能是奇异的)线性方程组 Ax=b (1.1) 其中A∈R^(n.n),b∈R(A),x∈R^n. 求解(1.1)的线性定常迭代可表为 X^(v+1)=Tx^v+g, (1.2) 其中T∈R^(n,n)称为(1.2)的迭代矩阵。一般。We consider polynomial acceleration methods for symmetric positive scmidcfinitc systems Ax = b based on basic iterative mehods which arc based on splittings of A = M-N,where M is not necessarily nonsingular.The Chcbyshcv semi-iterative algorithm and the preconditioning conjugate gradient method (generalized conjugate gradient method) arc established,their convergence and iteration error bounds arc studied.These arc then used to obtain the corrcspoding results when the basic iterative methods arc taken as block JOR and blocck SSOR.

关 键 词：对称正半定组 多项式加速法 迭代 

分 类 号：O24.16[理学—计算数学]