一种新的求解线性方程组的外推加速方法

A NEW EXTRAPOLATION ACCELERATION METHODS FOR LINEAR SYSTEMS

本文提出两种优化模型,通过在子空间{x^((k))…,x^((k-m))}上寻找最优解,建立了一种新的外推加速方法.讨论了该方法的收敛性和收敛速度.最后,通过三个数值实例展示了算法是可行的和有效的.

In this paper, we present two models of optimization and establish a new extrapolation acceleration method based on finding the optimal solution in subspace {x（k）,…,x（k=m）｜. We also discuss the convergence and convergent rate. Finally, we show that the new method is feasible and effective by three numerical examples.