非线性互补问题的改进超松弛迭代算法

Modified Overrelaxation Method for Nonlinear Complementarity Problem

运用松弛迭代算法与矩阵分裂理论,提出了求解非线性互补问题的改进超松弛迭代算法.这类算法设计了两个参数:第一个参数控制了迭代阵的谱半径,从而使算法收敛,适当选取第二个参数,加快了算法的收敛速度.在一定条件下证明了算法的全局收敛性.

In this paper,by using relaxation algorithm and matrix splitting,the authors first present a modified over-relaxation algorithm for nonlinear complementarity problem.The algorithm has two parameters,one ensures that the algorithm converge by controlling the spectral radius of the iteration matrix,the other one quickens the algorithm.And then,the global convergence theories of the algorithm are given.