Nonlinear Optimization Exam Sample
- Write the KT problem corresponding to the NLP problem:

- Given the NLP problem

Check if the constraint qualification holds at

- Take the NLP problem

Use the appropriate KT theorem to classify the point
State clearly the requirements of the theorem and test if they apply on this
problem.
- The minimizer of the single-variable function
is
being sought. It is known that the function has its minimizer in the interval
[-1, 0]. Reduce the interval of uncertainty by applying the Fibonacci method
with n = 4 and report the limits of the final interval.
- Consider the unconstrained function of two variables

Which of the following two directions is a descent direction at
?

- We seek to find the minimizer of the function

If
, find
using the steepest descent method.
- We want to use the conjugate gradient method to solve the unconstrained NLP
problem

Starting from
, it has
been calculated
,
.
Starting from this information, find
. Is
an optimal
solution?
- Take the NLP problem

- Write the composite function consisting of the exterior penalty function.
- Without carrying the actual calculations, indicate how the starting solution should
be selected if the exterior penalty function will be used to solve the
problem?
- What would be a good starting value for the penalty parameter?
- DFP quasi-Newton method will be used to solve the unconstrained problem
 Given:

calculate the direction
Maintain 3 decimal places.
- Use linear approximation to approximate the NLP problem

by a linear model at
.
- We want to solve the NLP problem
 Let
.
If the starting basic variables are chosen to be x1, x3,
and x5, use the reduced gradient method to find x1.
Note: use the method of augmented matrix to find the inverse of a matrix:
 Dr Muhammad Al-Salamah
Industrial Engineering
Systems Engineering Department
King Fahd University
Saudi Arabia
2008
|
|