Accelerated Relaxation Modulus-Based Matrix Splitting Iteration Method for Linear Complementarity Problems

Abstract

For solving large sparse linear complementarity problems effectively, by utilizing the parametric method, the acceleration strategy and the relaxation technique to the modulus-based matrix splitting (MMS) iteration method, we develop the accelerated relaxation MMS (ARMMS) iteration method, which generalizes the generalized accelerated MMS (GAMMS) and the relaxation MMS (RMMS) ones proposed recently. Meanwhile, it is proved that the ARMMS iteration method is convergent under proper restrictions when the system matrix is a positive definite matrix or an $$H_{+}$$ -matrix. In addition, we also study the convergence properties of the ARM accelerated overrelaxation (ARMAOR) method. Numerical experiments show that the proposed iteration method is efficient, and it outperforms some existing ones with suitable choice of the parameter and matrix splitting.