项目名称: 求解一类大规模稀疏线性矩阵方程的高效算法研究
项目编号: No.11501272
项目类型: 青年科学基金项目
立项/批准年度: 2016
项目学科: 数理科学和化学
项目作者: 李旭
作者单位: 兰州理工大学
项目金额: 18万元
中文摘要: 大规模稀疏线性矩阵方程存在于科学计算与工程应用的许多领域,设计高效的求解这种矩阵方程的数值算法对于解决这些领域中的实际问题起着非常重要的作用。但是到目前为止,真正适合这类矩阵方程求解的数值迭代方法很少,而对于求解系数矩阵是复矩阵的矩阵方程的研究则更为少见。为了有效解决矩阵方程的数值求解问题,本项目将在借鉴线性矩阵方程的HSS迭代方法设计经验的基础上,通过分析系数矩阵的特点和性质,研究适用于求解矩阵方程的高效数值迭代算法。相信通过本项目的研究,可以为矩阵方程的求解提供高效数值算法,进而为相关领域实际问题的快速有效解决提供算法保障。
中文关键词: 线性矩阵方程;迭代法;收敛性分析;Hermitian和反Hermitian分裂;谱半径
英文摘要: Large sparse linear matrix equations arise in a wide variety of scientific computing and engineering applications. Designing efficient algorithms for this kind of matrix equations is a very important task for the settlement of the practical problems. Till now, few algorithms can be used to solve this kind of matrix equations. The iteration methods used for solving the matrix equations with complex coefficient matrix are even rarer. In order to solve the matrix equations efficiently, this project will study high efficient numerical algorithms by referencing the designing experience of the HSS iteration method for linear matrix equations and analyzing the characteristics and properties of the coefficient matrix. We believe that, through the research of this project, high efficient iteration methods will be proposed for the linear matrix equations. Furthermore, the needs for the fast settlement of the associated practical problems will be satisfied.
英文关键词: linear matrix equation;iteration method;convergence analysis;Hermitian and skew-Hermitian splitting;spectral radius