摘要
将基于稀疏矩阵通用的按行存储方法的广义最小化残差算法应用于大型河网数值计算中,使大型河网的节点编号更具任意性,只需保证节点编号前后的一致性,而无需优化节点编码.实例计算结果表明,该算法精确、高效.
The general minimum residual algorithm (GMRES) based on the storage scheme by rows for an unstructured sparse matrix is applied to the numerical calculation of a large-scale river network. This algorithm helps the node number of the large-scale river network to be more arbitrary; the node number is not to be optimized and is only to keep consistent. Calculation results of an example show that the algorithm is accurate and efficient.
出处
《水利水运工程学报》
CSCD
北大核心
2003年第4期45-48,共4页
Hydro-Science and Engineering
关键词
非恒定流
河网
GMRES算法
稀疏矩阵
数值计算
unsteady flow
river network
GMRES algorithm
sparse matrix
numerical calculation
