[关闭]
@lyc102 2016-05-23T05:56:10.000000Z 字数 5143 阅读 1682

Operator Splitting

math


Problem

The function can further dependen on the spatial variable and could a be linear/nonlinear operator involving partial differential operator in space, i.e. (1) could be a PDE not just ODE. Indeed the operator splitting method is mostly used as PDE not ODE solvers.

Forward vs Backward Euler Methods

Going from to , we can use either

Basic Idea of Operator Splitting Methods

Write with ``simpler" operators and . Then evolve them using implicit and explicit scheme alternatively.

The Peaceman-Rachford method
From to , we set .

The Douglas-Rachford method

Write out detailed algorithm.

Example: C-N and Euler Methods
or
- P-R will give two versions of C-N
- D-R will give two versions of Euler's method

Example: ADI (Alternating Direction Implicit) Method
.


Each step solve a block tri-diagonal matrix which is much easier than solving .

Stability

Consider the simple linear problem

with a SPD operator . After a time discretization, we could write generitically .

Stablity: for some norm , we have

With such stability, we could obtain the error estimate
where is the truncation error.

The stability is equivalent to study the norm for some approriate norms (e.g. , or -norm). As is SPD and is a rational polynomail, it is reduced to study the function , where is an eigenvalue of and .

Consider the inequality . If it holds for all , we call it unconditionally stable. Otherwise the upper bound put a restriction on the step size which is usually in the form .

For general operator , we could extend the domain of from the real line to the complex plane and consider range of , the real part of .

Remark In ODE solvers, the eqn is in the form which is a sign difference with our setting. For example, the unconditional stable means for all with not in our setting.

Stronger Stability

The exact solution of (*) satisfy no matter what the initial condition is. After a time discretization, we would like to preserve the decay property. To do so, we need a stronger stability (also known as A-stable or L-stable) result

It might happen which is acceptable. As corresponds to small eigenvalues for which the component decay with a smaller rate (the graph is relative flat) and thus the contraction rate close to is acceptable for , i.e. for small eigenvalues of (fast transition vs slow transition). For large eigenvalues of , the exact decay rate is large. Then close is not accurate.

Consider the case of but . Then to capture the decay property in the discretization, is still needed. When is large, a small step size is needed to have the correct decay property even the solution is smooth after a fast transition. This difficulty is known as stiffness.

Consider

The exact solution
where is the so-called steady-state solution (independent of ). The solution will converge to as .

Stronger stable scheme will be also more suitable for computing the steady-state solution.

Crank-Nicolson method is unconditional stable and second order time discretization but it is not strong stability. The polynomial as .

Backward Euler is strong stable as but only first order.

A second order and strong stable scheme is BDF_2: suppose are known, we obtain by solving

Accuracy

Formally this is done by comparing Taylor series. First


Then we expand at . If , then we say the scheme of order . The power is reduced by one since we consider the finite difference approximation to partitial derive of and cancel one in the denominator.

Example

A -method

Now we set two inter middle points in one time intervale. Given , we set and . Again we split and iterative:

What is the optimal choice of ? How do we split ? We study a simplified problem and splitting scheme: is linear and .

Write out and study the stability and accuracy.

添加新批注
在作者公开此批注前,只有你和作者可见。
回复批注