严格次对角占优线性方程组迭代法的收敛性分析

Convergence analysis of iterative methods for strictly sub-diagonally dominant linear equations

Jacobi迭代法、Guass-Seidel迭代法和SOR迭代法是求解线性方程组的常用迭代方法.本文证明了系数矩阵严格次对角占优时,Jacobi迭代法、Guass-Seidel迭代法和SOR迭代法均收敛,并给出了相应的误差估计.通过比较三种迭代法的误差上界,指明Guass-Seidel迭代法的误差上界最小.

The Jacobi iterative method, Guass-Seidel iterative method, and SOR iterative method are commonly used in solving linear equations. When the coefficient matrix of a system of linear equations is strictly sub-diagonally dominant, we demonstrate that the Jacobi, Guass-Seider, and SOR iterative methods are all convergent. By comparing the upper bounds of error for the three iterative methods, we show that the upper bound of error for the Guass-Seidel iterative method is minimal.