Three methods for the parallel solution of a large, sparse system of linear equations by multiprocessors

摘要

In this paper three methods for the parallel solution of large-scale, sparse, linear equation systems are presented and compared. Both, factorization of the matrix and forward backward. Processes have been considered. A multiprocessor architecture has been simulated and results are presented corresponding to eight examples taken from power systems field, ranging from 117 to 660 equations.