Convergence study on numerical methods solving Hyperbolic PDEs (双曲型pde 的数值方法稳定性研究)
来源:互联网 发布:淘宝直通车卡首屏 编辑:程序博客网 时间:2024/05/04 15:36
重读大部头书《 Finite Volume Methods for Hyperbolic Problems》。现在知识碎片太多,重读大书能助提纲携领。
* Hyperbolicalicity of linear systems
q_t + A q_x =0
matrix A is diagonalizable with real eigenvalues, do diagonalization --> w_t + D w_x = 0, where D is diagonal matrix, now the system is decoupled. -->
w_t ^ p + \lamba w_x^p = 0
--> solutions of original system now can be consisted of a linear combination of m "waves" traveling at characteristic speeds ( \lamba ).
Define characteristic curves X(t) = x_0 + \lamba^p t , along which the information propagates
*Domain of dependence & Range of influence
- Let <X,T> be fixed point in space-time, then the solution w(X,T) depends only on some particular points ( X - \lamba^p T) from the initial line. (p = 1, ... m), this is called the domain of dependence of <X,T>. on the contrary way, think about which particles will be influenced by the current <X,T> will give the range of influence as < X + \lamba^p T> .
- For hyperbolic PDE, domain of dependence is always bounded, resulting form the fact that information propagates at finite speed. In design of numerical method means that the explicit methods will be efficiently. While for parabolic PDE, domain of dependence is the entire initial line, so implicit methods is needed.
- For discontinuous in initial data, only propagate along characteristics for a linear system. meaning, the discontinuous in initial values will extension only along characteristic curves.
* General Consideration about Convergence and Stability
- consistent, numerical solution should approximate well locally | y_{i+1} - y_i | <= L (x_{i+1} - x_i )
- stable in some sense (L^2 norm usually), the small errors made in each time step do not grow too fast in later time steps.
- consistence + stability == convergence
* CFL conditional number
Define: a numerical method can be convergent only if its numerical domain of dependence contains the true domain of dependence of the original PDE.
Define: domain of dependence of a numerical approximation
- let dt/dx = r fixed, for point <X, T> in space-time , the domain of dependence lie in the initial line as
0 - T = r ( x- X)
0 - T = -r (x - X)
namely, [X - T/r, X + T/r]
the domain of depend of a original PDE is as above , D(X, T) = { X - \lamba^p T ; p = 1, 2, .... m }
so CFL requires X - T/r <= X - \lamba^p T <= X - T/r
define Courant Number \mu = dt/dx max| \lamba^p| <= 1
CFL is necessary for convergence, another stability analysis methods is by Von Neuman, by define amplification factor, represent each frequency in the solution is amplified in advancing the solution one time step.
Von Neuman 方法,在稀疏矩阵迭代方法的收敛性分析中也有应用,既迭代矩阵的谱半径不超过1。 下次再复习下parabolic pde 的数值解性质。 这样把基础理论过一遍。
- Convergence study on numerical methods solving Hyperbolic PDEs (双曲型pde 的数值方法稳定性研究)
- 数值优化(Numerical Optimization)学习系列-惩罚和增广拉格朗日方法(Augmented Lagrangian Methods)
- Solving PDEs in Minutes
- 数值优化(Numerical Optimization)学习系列-信赖域方法
- Operator Splitting Methods in PDEs
- homework of numerical solution of PDEs
- 【索引】Numerical Methods
- 【索引】Numerical Methods::Examples
- 数值优化(Numerical Optimization)学习系列-线搜索方法(LineSearch)
- 数值优化(Numerical Optimization)学习系列-共轭梯度方法(Conjugate Gradient)
- 数值优化(Numerical Optimization)学习系列-拟牛顿方法(Quasi-Newton)
- 使用PDE的方法进行图像修复
- 调用nil对象的方法(Calling Methods on Nil)
- 调用nil对象的方法(Calling Methods on Nil)
- 数值优化(Numerical Optimization)学习系列(二)-无约束优化的基础
- Applied Numerical Methods Using MATLAB
- Numerical Analysis : Several Interpolation Methods
- 【索引】Numerical Methods::Exercises: Beginner
- 区间的不同实现方式
- LEETCODE CLASSIFICATION
- Threaded Binary Tree
- 如何为iPhone
- 【最具价值CTO评选秀】出门问问CTO雷欣:从技术极客到核心管理的秘密
- Convergence study on numerical methods solving Hyperbolic PDEs (双曲型pde 的数值方法稳定性研究)
- FSD是什么意思呀
- 【最具价值CTO评选秀】出门问问CTO雷欣:从技术极客到核心管理的秘密
- sparkSQL1.1入门之三:sparkSQL组件之解析
- FSD是什么意思呀?
- win7组播问题
- iOS8.1完美越狱
- [LeetCode] Add two numbers
- visual studio 2013 update 4 正式版出来了和visual studio 2015 预览版