A new version of KSOR method with lower number of iterations and lower spectral radius

In this paper, we propose a new method for solving linear algebraic systems based on the KSOR method. The new approach introduces a new coefficient in the assumed updated section from the KSOR method. It obtains the same solution as KSOR for a given linear algebraic system, but in order to get this,...

Full description

Saved in:
Bibliographic Details
Published in:Soft computing (Berlin, Germany) Germany), 2019-11, Vol.23 (22), p.11729-11736
Main Authors: Constantinescu, R., Poenaru, R. C., Pop, F., Popescu, P. G.
Format: Article
Language:English
Subjects:
Artificial Intelligence
Complexity
Computational Intelligence
Control
Eigenvalues
Engineering
Linear algebra
Mathematical Logic and Foundations
Mechatronics
Methodologies and Application
Methods
Robotics
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In this paper, we propose a new method for solving linear algebraic systems based on the KSOR method. The new approach introduces a new coefficient in the assumed updated section from the KSOR method. It obtains the same solution as KSOR for a given linear algebraic system, but in order to get this, we have to modify the diagonal of the coefficient matrix of the initial system. The new method reaches the solution in fewer steps than KSOR on certain examples, but we have an additional step for the modification mentioned above with O ( m ) complexity, while a normal step of iteration has O ( m 2 ) complexity, so overall our method has better performances than KSOR method. Furthermore, on certain examples, the spectral radius of the iteration matrix of the new method is strictly less than the one of KSOR and MKSOR methods.
ISSN:1432-7643
1433-7479
DOI:10.1007/s00500-018-03725-2