Numerical Methods Qualification Exam Problems and Solutions (University of Maryland)/Jan07 667

From Wikibooks, open books for an open world
Jump to navigation Jump to search

Problem 4[edit | edit source]

Let , suppose and assume is nonsingular . Consider the following iteration

Problem 4a[edit | edit source]

Derive the following error equation for

Solution 4a[edit | edit source]

Note the following identity


The error is given by

Problem 4b[edit | edit source]

Let be a fixed matrix. Find conditions on B that guarentee local convergence. What rate of convergence do you expect and why?

Solution 4b[edit | edit source]

Assume is invertible, is bounded, and is Lipschitz.


This implies local superlinear convergence.

Problem 4c[edit | edit source]

Find sufficient conditions on for the convergence to be superlinear. What choice of corresponds to the Newton method and what rate of convergence do you expect?

Solution 4c[edit | edit source]

Super linear convergence[edit | edit source]

as

Find condition for super linear convergence[edit | edit source]

From part(b)



Since , if



as , we have super linear convergence i.e.


Problem 5[edit | edit source]

Let be uniformly Lipschitz with respect to . Let be the solution to the initial value problem : . Consider the trapezoid method

.

Problem 5a[edit | edit source]

Find a condition on the stepsize that ensures (1) can be solved uniquely for .

Solution 5a[edit | edit source]

The implicit method can be viewed as a fix point iteration:



We want


which implies


Problem 5b[edit | edit source]

Define a local truncation error and estimate it. Examine the additional regularity of needed for this estimate.

Solution 5b[edit | edit source]

Re-writing (1) and replacing we have a formula for consistency of order p:



For uniform stepsize h

Expanding in Taylor Series about gives

Problem 5c[edit | edit source]

Prove a global error estimate for (1)

Solution 5c[edit | edit source]

Problem 6[edit | edit source]

Consider the 2-point boundary value problem

,

with constants and . Let be a uniform partition of [0,1] with meshsize .

Problem 6a[edit | edit source]

Use centered finite differences to discretize (2). Write the system as

and identify . Prove that A is nonsingular.

Solution 6a[edit | edit source]

Using Taylor Expansions, we can approximate the second derivative as follows



We can eliminate two equations from the n+2 equations by substituting the initial conditions into the equations for and respectively.


We then have the system



is nonsingular since it is diagonally dominant.

Problem 6b[edit | edit source]

Define truncation error and derive a bound for this method in terms of . State without proof an error estimate of the form

and specify the order s.

Solution 6b[edit | edit source]

Local Truncation Error[edit | edit source]

Bound for Local Truncation Error[edit | edit source]

Derive Bound for Max Error[edit | edit source]

Let , , and is the local truncation error.


Then



Subtracting the two last equations gives



Hence,


, that is the error has order 2.

Problem 6c[edit | edit source]

Prove the following discrete monotonicity property: If is the solution corresponding to a forcing , for , and then componentwise.

Solution 6c[edit | edit source]

Note that is a matrix and hence the discrete maximum principle applies. (See January 05 667 test for definition of matrix)


Discrete Maximum Principle[edit | edit source]


If , then .


Specifically let , then which implies