作成者 |
|
|
|
|
本文言語 |
|
出版者 |
|
|
発行日 |
|
収録物名 |
|
巻 |
|
号 |
|
開始ページ |
|
終了ページ |
|
出版タイプ |
|
アクセス権 |
|
JaLC DOI |
|
関連DOI |
|
|
|
関連URI |
|
|
|
関連情報 |
|
|
|
概要 |
The conventional SOR (Successive Over-Relaxation) method started from the dissertation by D. Young in 1950. Though the SOR method has been often utilized for the solution of problems which occur from ...a rich variety of applications, the SOR method has a certain issue that the SOR method greatly depends on spectrum of iteration matrix. As a consequence, convergence of the SOR method is not robust. In this article, we extend IDR (Induced Dimension Reduction) Theorem to designing of the residual of the SOR method, and accelerate its robustness of convergence. Numerical experiments indicate significantly improved residual compared with that of the conventional SOR method.続きを見る
|