Parallel chaotic multisplitting iterative methods for the large sparse linear complementarity problem (Q2732180)

A linear complementarity problem is considered: Find \(z\in\mathbb{R}^n\) such that NEWLINE\[NEWLINEMz+ q\geq 0,\quad z^T(Mz+ q)= 0\quad\text{and }z\geq 0.NEWLINE\]NEWLINE To solve such a problem by means of parallel computing some multisplitting iterative methods are known. In this paper a parallel chaotic multisplitting method for solving large sparse linear complementarity problems is presented.NEWLINENEWLINENEWLINEConvergence properties are discussed under certain assumptions. Some applicable relaxed variants of the presented algorithm and their convergence properties are investigated. There are numerical examples, the numerical results show high parallel efficiency.