DocumentCode :
1541597
Title :
On the convergence rate improvement of ICCG solver on the FE mesh
Author :
Lee, Hyang-Beom ; Jung, Hyun-Kyo ; Hahn, Song-Yop ; Choi, Kyung ; Kim, Hyoung-Joong
Author_Institution :
Dept. of Electr. Eng., Kunsan Nat. Univ., South Korea
Volume :
33
Issue :
2
fYear :
1997
fDate :
3/1/1997 12:00:00 AM
Firstpage :
1760
Lastpage :
1763
Abstract :
A nested dissection ordering algorithm is adopted into the ICCG (incomplete Cholesky factorization for conjugate gradient) iterative solver, which can improve the convergence rate of the ICCG method in some cases. The ordering effects on the FE meshes are investigated with the convergence rate variation. It turns out that the ordering in the FE meshes has little effect against the acceleration factor in practical use. The method is applied to induction motors
Keywords :
conjugate gradient methods; convergence of numerical methods; induction motors; matrix decomposition; mesh generation; FE mesh; ICCG solver; acceleration factor; convergence rate improvement; convergence rate variation; incomplete Cholesky factorization for conjugate gradient; induction motors; iterative solver; nested dissection ordering algorithm; Acceleration; Bandwidth; Character generation; Convergence; Finite element methods; Iron; Iterative algorithms; Iterative methods; Parallel processing; Sparse matrices;
fLanguage :
English
Journal_Title :
Magnetics, IEEE Transactions on
Publisher :
ieee
ISSN :
0018-9464
Type :
jour
DOI :
10.1109/20.582614
Filename :
582614
Link To Document :
بازگشت