Optimization of the relaxation parameter for S. S. O. R. and A. D. I. preconditioning (Q5961060)

The authors present and compare three different approaches for an optimization of the relaxation parameters both for symmetric successive overrelaxation (SSOR) and alternating directions implicit (ADI) basic iteration and the preconditioned conjugate gradient method. A description of the SSOR and ADI basic iterative methods viewed as a relaxation algorithm and as a preconditioned Richardson's type algorithm can be found in this paper. The detailed link between estimates of the spectral radius of the iteration matrix and the condition number of the preconditioning matrix, which is given in the paper, allows to choose the best approach in order to obtain the optimal relaxation parameter and the corresponding optimal estimates in different situations.