Sei sulla pagina 1di 3

exe2

COMP 567

1. (5) Show that the dual problem of

max

s.t.

b

T

y 0

T

y

y

A

c

is the following problem

min

s.t.

T

c

Ax

x

x 0

b

2. Consider the following linear programming problem

min

s.t.

n

c

j = 1

j

x

j

n

a

j = 1

ij

l

j

x

x

j

j

=

b

i

u

j

i

=

1,2,

,

m

 

j

= 1,

2,

,

n

.

a) (5) State the dual of this problem.

b) (20) Show that if x is feasible for this problem, and if there exists a vector

of multipliers

π =

c

c

j

j

+

+

[

π

1

,

π

2

,

m

i = 1

m

i = 1

π

i

π

i

a

a

ij

ij

>

<

,

0

0

π

m

]

T such that

x

x

j

j

=

=

l

j

u

j

then x is an optimal solution of this problem.

2.

(20) Suppose that x* and y* are optimal solutions of the following pair of primal-dual problems :

min

c

T

x

s.t.

Ax

b

 

x

0

max

s.t.

T

b

A

y

T

y

c

y 0.

Suppose also that x' is an optimal solution of the following problem

T

Show that c x

′ ≥

c

T

min

s.t.

x *

+

k

T

c

T

x

Ax

b

+

x

0.

y *.

k

4. Consider the following problem

(Primal)

max

s.t.

z

=

x

1

x

1

2

,

x

1

x

2

x

2

4

1

x

0

2

a) (5) State the dual problem, and solve it by observation.

b) (5) Use the complementary slackness conditions and the optimal solution of the dual problem to determine an optimal solution of the primal problem.

c) (5) Consider the coefficient c 1 of x 1 in the objective function of the primal problem, determine the values of c 1 for which the dual problem does not have a feasible solution? For these values of c 1 , what is happening to the primal problem?

5. (15) Use the dual simplex method to solve the problem using the slack variables as the basic variables at the first iteration :

min

s.t.

3 x

1

+

x

2

+

x x

1 2

+

2

x

3

3

x x

1 2

x

1

x

1

,

x

x

2

2

2

x

3

,

+

4

x

3

x

3

0.

8

6

2