In solving the large-scale sparse complex linear equation group, a lot of CPU operational time and computer memory is consumed to access zero elements of sparse coefficient matrix. As an effective solution, a fully- s...
详细信息
In solving the large-scale sparse complex linear equation group, a lot of CPU operational time and computer memory is consumed to access zero elements of sparse coefficient matrix. As an effective solution, a fully- sparse storing scheme is proposed to store only nonzero el- ements of symmetrical part by chain pattern. For some ill- conditioned coefficient matrixes, iterative solution meth- ods may incur such problems as slow convergence and even failure of convergence. Fortunately, some valid precondi- tioning techniques can improve the convergence by reduc- ing condition number of ill-conditioned matrix. Based on a real incomplete Cholesky factorization preconditioner, we develop a fast convergent preconditioned Bi-conjugate gra- dient method (BCG) to solve the large-scale sparse com- plex linear equation group. Numerical experiments show that the new incomplete Cholesky factorization precondi- tioner accelerates the convergence and preconditioned bi- conjugate gradient method is available for the large-scale complex linear equation group.
暂无评论