Submitted by Automa??o e Estat?stica (sst@bczm.ufrn.br) on 2017-02-13T20:18:37Z No. of bitstreams: 1 JosimaraTatianeDaSilva_DISSERT.pdf: 1557682 bytes, checksum: fac59260c784cbb83579953ae2c457f9 (MD5) / Approved for entry into archive by Arlan Eloi Leite Silva (eloihistoriador@yahoo.com.br) on 2017-02-16T19:30:20Z (GMT) No. of bitstreams: 1 JosimaraTatianeDaSilva_DISSERT.pdf: 1557682 bytes, checksum: fac59260c784cbb83579953ae2c457f9 (MD5) / Made available in DSpace on 2017-02-16T19:30:20Z (GMT). No. of bitstreams: 1 JosimaraTatianeDaSilva_DISSERT.pdf: 1557682 bytes, checksum: fac59260c784cbb83579953ae2c457f9 (MD5) Previous issue date: 2016-07-19 / Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior (CAPES) / Este trabalho tem por objetivo investigar o comportamento de converg?ncia do m?todo GMRES (Generalized Minimal RESidual) e sua vers?o GMRES(m), sem e com precondicionador ILU(0) aplicado ? sistemas lineares n?o sim?tricos esparsos. Nosso interesse principal ? verificar se o comportamento destes algoritmos pode ser influenciado pela estrutura das matrizes consideradas, em particular, as Z-matrizes e a influ?ncia da escolha do grau de esparsidade. Entre os par?metros observados, concentramos no raio espectral dessas matrizes, tanto como a norma do res?duo relativo obtido por estes algoritmos. / This study aims to investigate the convergence behavior of the GMRES (Generalized Minimal Residual) method and its version GMRES(m), without and with preconditioner ILU(0) applied to sparse non-symmetric linear systems. Our main interest is to see if the behavior of these algorithms can be influenced by the structure of the matrices considered, in particular, the Z-matrices. Furthermore, the influence of the choice of the degree of sparsity. Among the observed parameters, we focus on the spectral radius of these matrices, as well as the relative residual norm obtained by these algorithms.