首页 | 本学科首页   官方微博 | 高级检索  
     检索      

Krylov子空间投影法及其在油藏数值模拟中的应用
引用本文:刘晓明,卢志明,刘宇陆.Krylov子空间投影法及其在油藏数值模拟中的应用[J].应用数学和力学,2000,21(6):551-560.
作者姓名:刘晓明  卢志明  刘宇陆
作者单位:上海市应用数学和力学研究所, 上海大学, 上海 200072
基金项目:国家自然科学基金!资助项目 (19872 0 4 3)
摘    要:Krylov子空间投影法是一类非常有效的大型线性代数方程组解法,随着左右空间Lm、Km的不同选取可以得到许多人们熟知的方法.按矩阵Hm的不同类型,将Krylov子空间方法分成两大类,简要分析了这两类方法的优缺点及其最新进展.将目前最为可靠实用的广义最小余量法(GMRES)应用于油藏数值模拟计算问题,利用矩阵分块技术,采用块拟消去法(PE)对系数阵进行预处理.计算结果表明本文的预处理GMRES方法优于目前使用较多的预处理正交极小化ORTHMIN方法,最后还讨论了投影类方法的局限和今后的可能发展方向.

关 键 词:Krylov子空间投影法    块拟消去法    油藏数值模拟
收稿时间:1999-03-15

Krylov Subspace Projection Method and Its Application on Oil Reservoir Simulation
Liu Xiaoming,Lu Zhiming,Liu Yulu.Krylov Subspace Projection Method and Its Application on Oil Reservoir Simulation[J].Applied Mathematics and Mechanics,2000,21(6):551-560.
Authors:Liu Xiaoming  Lu Zhiming  Liu Yulu
Institution:Shanghai Institute of Applied Mathematics and Mechanics, Shanghai University, Shanghai 200072, P R China
Abstract:Krylov subspace projection methods are known to be highly efficient for solving large linear systems.Many different versions arise from different choices to the left and right subspaces.These methods were classified into two groups in terms of the different forms of matrix Hm,the main properties in applications and the new versions of these two types of methods were briefly reviewed,then one of the most efficient versions,GMRES method was applied to oil reservoir simulation.The block Pseudo-Elinimation method was used to generate the preconditioned matrix.Numerical results show much better performance of this preconditioned techniques and the GMRES method than that of preconditioned ORTHMIN method,which is now in use in oil reservoir simulation.Finally,some limitations of Krylov subspace methods and some potential improvements to this type of methods are furtherly presented.
Keywords:Krylov subspace methods  block PE method  numerical oil reservoir simulation
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《应用数学和力学》浏览原始摘要信息
点击此处可从《应用数学和力学》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号