逐次超松弛法因子的选取

The Selection of Factors about Successive Overrelaxation Method

文章在系数矩阵A满足对称正定的情况下给出了一类解大型稀疏线性系统Ax=b的最新方法,即渐近最优超松弛迭代法,避免了传统选择最佳松弛因子带来的不便,并通过理论性证明此算法收敛于Ax=b的解或近似解.

We present a class of asymptotically optimal successive overrelaxation methods for solvingthe large sparse system of linear equations. Avoiding inconvenience for the traditional choose of the bestrelaxation factor. And we prove the solution or approximate solution of the al- gorithm convergence Ax=b as thecoefficient matrix is a symmetric positive definite matrix.