Authors: Guo-Yan Meng; Qing-Shan Zhao; Yulan Hu
Addresses: Department of Mathematics, Xinzhou Normal University, Xinzhou, Shanxi Province, China ' Department of Computer, Xinzhou Normal University, Xinzhou, Shanxi Province, China ' Department of Computer, Xinzhou Normal University, Xinzhou, Shanxi Province, China
Abstract: In the Hermitian and skew-Hermitian splitting (HSS) iteration method, the determination of the optimal parameter is a tough task when solving a large sparse non-Hermitian positive definite linear systems. In this paper, we present the variable HSS iteration with the non-fixed positive constant. For obtaining the approximate optimal parameters of this method, we solve the minimising the optimisation model based on the bacterial foraging optimisation (BFO) algorithm. Numerical experiments have shown that the new strategy is feasible and effective than the HSS iteration method.
Keywords: variable HSS iteration; optimal parameters; bacterial foraging optimisation; BFO; linear systems; Hermitian and skew-Hermitian splitting.
International Journal of Computing Science and Mathematics, 2015 Vol.6 No.5, pp.471 - 479
Received: 04 Jun 2015
Accepted: 10 Jul 2015
Published online: 10 Nov 2015 *