SSG 811 Solution To Assignment

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 15

QUESTION 1

1. Given the linear programme below:

Max.z = 4x + y

Subject to:

x + y ≤ 50

3x + y ≤ 90

x ≥ 0, y ≥ 0

a. Solve the problem using graphical method

90

60

(0, 50)

(20, 30)
30

(30, 0)
x
(0, 0)
10 20 30 40 50

z = 4x + y

At (0,0), z = 4 × 0 + 0 = 0

At (0, 50), z = 4 × 0 + 50 = 50

At (20, 30), z = 4 × 20 + 30 = 110

At (30, 0), z = 4 × 30 + 0 = 120

The optimal value is z = 120


Occurring at x = 20, y = 30

b. Using graphical method, determine the dual or shadow prices due each resource available at
each constrain resource.

Consider unit change of RHS

90 G

60

A(0, 50)

H(19.5, 30.5) x + y ≤ 51
(20, 30)
30
B

F
C(30, 0)
x
O(0, 0)
10 20 30 40 50

Constraint 1 becomes

x + y ≤ 50 + 1 = 51 i.e. x + y ≤ 51

From this we can study the shift in the optimal solution point

New intersect solves

x + y = 51

3x + y = 90

39 61
It gives x = , y=
2 2

39 61
New intersection value of z = 4 × + = 108.5
2 2

Hence,
zE − zB 108.5 − 110,0
Rate of Revenue change due to increase of 1 unit capacity = = = −
Change in capacity 51 − 50
N1.5 / unit (Dual Price/ Shadow Price or unit worth of resource)

y
G
90

3x + y ≤ 91

60

A (0, 50)

(20, 30)
30 B E(20.5, 29.5)

F
C (30, 0)
x
O (0, 0)
10 20 30 40 50

Constraint 2 becomes

3x + y ≤ 90 + 1 = 91 i.e. 3x + y ≤ 91

From this we can study the shift in the optimal solution point

New intersect solves

x + y = 50

3x + y = 91

41 59
It gives x = , y=
2 2

41 59
New optimal value of z = 4 × + = 111.5
2 2

Hence,
zE − zC 111.5 − 110,0
Rate of Revenue change due to increase of 1 unit capacity = = =
Change in capacity 91 − 90
N1.5 / unit (Dual Price/ Shadow Price or unit worth of resource)

c. Determine the ranges of value for which changes D1 and D2 in resource available for each dual or
shadow price holds for the dual prices to remain valid.

From the graphs,

Constraint 1

Minimum operation 1 unit capacity (at C = (30, 0)) = 30 + 0 = 30 unt (from constraint equation 1)

Maximum operation 1 time capacity (at G = (0, 90)) = 0 + 90 = 90 units (from constraint equation 1)

Thus, the dual price of –N1,5 unit will remain valid for the feasibility range:

30 units ≤ Operation 1 capacity ≤ 90 units.

(19.5,
Changes outside this range will produce 30.5)
a different dual price.

For Constraint 2

Minimum operation 1 unit capacity (at F = (0, 50)) = 3 × 0 + 50 = 50 unt (from constraint equation 2)

Maximum operation 1 time capacity (at A = (50, 0)) = 3 × 50 + 0 = 150 units (from constraint equation 2)

Thus, the dual price of –N1,5 unit will remain valid for the feasibility range:

50 units ≤ Operation 2 capacity ≤ 150 units.

Changes outside this range will produce a different dual price.

d. Determine the optimality range of each of the unit costs (objective coefficients) in the problem for
which the optimal solution remains valid

QUESTION 2
2. A company produces three products A, B, and C whose unit revenue yield are 12, 3 and 1 (in hundreds
of thousands of Naira) respectively. Three raw materials, M1, M2 and M3 are used to produce three
products. Consumption rates of material M1 for each unit of products A, B and C produced are 10, 2 and
1 kg respectively. For M2 the consumption rates are 7, 3 and 2 kilogrammes respectively while for
material M3 the consumption rates are 2, 4 and 1 kilogrammes respectively. The maximum availability of
raw materials M1, M2 and M3 are 100, 77 and 80 kilogrammes respectively.

a. Formulate the problem as a linear programme.

Let the quantity of product A be x1

Let the quantity of product B be x2

Let the quantity of product C be x3

From the problem description,

Revenue (in hundreds of thousands of Naira) yielded will be

z = 12x1 + 3x2 + x3 (since unit revenue yields of products A, B and C are 12, 3 and 1 hundreds of
thousands respectively)

This is to be maximized. Hence the objective of the problem is to

Maximize z = 12x1 + 3x2 + x3

Usage of raw materials M!, M2 and M3

Material M1: Use 10, 2 and 1 kg for every unit of products A, B, and C respectively. Maximum
availability is 100 kg. Hence,

10x1 + 2x2 + x3 ≤ 100

Material M2: Use 7, 3 and 2 kg for every unit of products A, B, and C respectively. Maximum availability
is 77 kg. Hence,

7x1 + 3x2 + 2x3 ≤ 77

Material M3: Use 2, 4 and 1 kg for every unit of products A, B, and C respectively. Maximum availability
is 80 kg. Hence,

2x1 + 4x2 + x3 ≤ 80

Thus the three associated constraints are formulated.

Non-Negativity Requirements

Quantities x1, x2, x3 are required to be zero or positive (i.e. non-negative) Thus:
x1 ≥ 0, x2 ≥ 0, x3 ≥ 0

Altogether, the whole formulated Linear programme becomes

Maximize z = 12x1 + 3x2 + x3

Subject to:

10x1 + 2x2 + x3 ≤ 100

7x1 + 3x2 + 2x3 ≤ 77

2x1 + 4x2 + x3 ≤ 80

x1 ≥ 0, x2 ≥ 0, x3 ≥ 0

b. Solve the problem using Simplex method.

Step 1: Standardize the programme

We introduce 3 slack variables s1, s2 and s3 to the constraints respectively.

Maximize z − 12x1 − 3x2 − x3 = 0

Subject to:

10x1 + 2x2 + x3 + s1 = 100

7x1 + 3x2 + 2x3 + s2 = 77

2x1 + 4x2 + x3 + s3 = 80

x1 ≥ 0, x2 ≥ 0, x3 ≥ 0, s1 ≥ 0 , s2 ≥ 0, s3 ≥ 0

Step 2: Determine the Initial (Starting) Feasible solution

No of variables = 6

No of constraints = No of basic variables = 3

Therefore,

No of non-basic (zero) variables = No of variables – No of Basic Variables = 6 – 3 = 3

So, we are required to set 3 variables each to zero so as to obtain a feasible solution at every constraint
equation i.e.

By inspection, set x1 = 0 , x2 = 0, x3 = 0
This gives values of basic variables

s1 = 100

s2 = 77

s3 = 80

Step 3: Set up the initial Simplex table (Iteration 0):

Step k: perform Gauss Jordan, Apply Optimality Condition, Apply Feasibility Condition

Initial non-basic variables Initial Basic Variables


Iteration Basic Var. x1 x2 x3 s1 s2 s3 Solution
Ratio
z- -12 -3 -1 0 0 0 0
0
s1 10 2 1 1 0 0 100 10
x1 enters
s2 7 3 2 0 1 0 77 11
s1 leaves
s3 2 4 1 0 0 1 80 40
z- 0 - 3/5 1/5 1 1/5 0 0 120
1
x1 1 1/5 0 0 0 0 10 50
x2 enters
s2 0 1 3/5 1 2/7 - 5/7 1 0 7 4 3/8
s2 leaves
s3 0 3 3/5 4/5 - 1/5 0 1 60 16 2/3
z- 0 0 2/3 1 3/8 0 122 5/8
2 x1 1 0 -0 1/5 - 1/8 0 9 1/8
OPTIMAL x2 0 1 4/5 - 4/9 5/8 0 4 3/8
s3 0 0 -2 1/8 1 3/8 -2 ¼ 1 44 ¼

c. From the solution in (b) identify the dual or shadow prices of raw materials M1, M2 and M3.

From the Optimal Table, the shadow prices are the z-values of the slack variables at the optimal table
iteration.

Here,
Basic x1 x2 x3 s1 s2 s3 Solution
Z 0 0 2/3 1 3/8 0 122 5/8

Resource Slack Variable Optimal z-equation coefficient of slack variable Dual Prices

Material M1 s1 1 N1 / kg
Material M2 s2 3/8 N3/8/kg.
Material M3 s3 0 N0 / kg
Notice that s3 is basic at the optimal level has no dual price

Important Deductions

i. The zero dual price for material M3 means that there is no economic advantage in allocating
more material type M3 to the production.
ii. This result in (i) is in order and makes sense in that the result is already abundant as
evidenced by the slack variable associated with Material M3 is positive (= 44 1/4) in the
optimum solution.
3
iii. As for materials M1 and M2, one kilogramme increase will improve revenue by N1 and N
8
(in hundreds of thousands) respectively. When allocating additional material, material M1
may have higher priority because its dual price more than that of material M2.
iv. The computations above are for constraints with ≤ . When the constraints are of ≥ , the same
idea is applicable except that the dual price will assume the opposite sign of that associated
with ≤ constraint

d. Suppose the levels of availability of M1, M2 and M3 change by 20, 10 and 15 kilogrammes
respectively, do the current dual prices hold?

Let D1, D2, D3 be changes (positive or negative) in the availability of materials M1, M2 and M3
respectively. The model can be written as:

Maximize z = 12x1 + 3x2 + x3

Material M1: 10x1 + 2x2 + x3 ≤ 100 + D1

Material M2: 7x1 + 3x2 + 2x3 ≤ 77 + D2

Material M3: 2x1 + 4x2 + x3 ≤ 80 + D3

x1 ≥ 0, x2 ≥ 0, x3 ≥ 0
We need to recomputed the optimum Simplex tableau with the modified right hand side and from it derive
the condition that keeps the solution feasible. It is in these conditions that we substitute the given specific
vales of changes and see if the current solution remains optimal or not.

We don’t need to solve from scratch as there are obvious computation similarities that we can leverage
on.

The initial iteration of the solution can be arrayed as below:

Initial non-basic variables Initial Basic Variables

Iteration Basic Var. x1 x2 x3 s1 s2 s3 RHS D1 D2 D3

z- -12 -3 -1 0 0 0 0 0 0 0

0 s1 10 2 1 1 0 0 100 1 0 0
x1 enters
s2 7 3 2 0 1 0 77 0 1 0
s1 leaves

s3 2 4 1 0 0 1 80 0 0 1

The two shaded portions are similar from the onset.

Since it is the same computations that will be performed to optimal solution, the solutions of the portions
at the optimal iteration will also be the same.

Thus, the new optimal table will be

Initial non-basic Initial Basic


variables Variables

Basic
D2 D3
Iteration x1 x2 x3 s1 s2 s3 RHS D1
Var.

z- 0 0 2/3 1 3/8 0 122 5/8 1 3/8 0


x1 1 0 -0 1/5 - 1/8 0 9 1/8 1/5 - 1/8 0
2
OPTIMAL x2 0 1 4/5 - 4/9 5/8 0 4 3/8 - 4/9 5/8 0
s3 0 0 -2 1/8 1 3/8 -2 ¼ 1 44 ¼ 1 3/8 -2 ¼ 1
It gives the following optimal solutions

z = 122 5 / 8 + D1 + 3/8 D2

x1 = 9 1 / 8 + 1/5D1 − 1/8D2

3 5
x2 = 4 − 4/9D1 + D2
8 8

1 3 1
s3 = 44 + 1 D1 − 2 D2 + D3
4 8 4

This new z-value also confirm the dual prices earlier deduced.

The current solution remains feasible as long as all the variables are nonnegative, leading to the
feasibility Conditions:

1
x1 = 9 + 1/5D1 − 1/8D2 ≥ 0
8

3 5
x2 = 4 − 4/9D1 + D2 ≥ 0
8 8

1 3 1
s3 = 44 + 1 D1 − 2 D2 + D3 ≥ 0
4 8 4

Any simultaneous changes D1, D2 and D3 that satisfy these inequalities will keep the solution feasible. If
all the conditions are satisfied, then the new optimum solution can be found through direct substitution of
D1, D2, D3in the equations given above.

Here, we are given 20, 10 and 15

D1 = 20

D2 = 10

D3 = 15

Substituting,

1 7
x1 = 9 + 1/5 × 20 − 1/8 × 10 = 11 ≥ 0 (feasible)
8 8

3 5 53
x2 = 4 − 4/9 × 20 + × 10 = 1 ≥0 (feasible)
8 8 72

1 3 1 1
s3 = 44 + 1 × 20 − 2 × 10 + 15 = 64 ≥ 0 (feasible)
4 8 4 4
This shows that the solution obtained remains optimal when these changes are made on the right hand
side or in availability of materials M1. M2 and M3.

The new objective value is:

5 5 3
z = 122 + D1 + 3/8 D2 = 122 + 20 + × 10 = 146.375
8 8 8

OR

7 53 1 1140 375 257


z = 12x1 + 3x2 + x3 = 12 × 11 + 3 × 1 + 64 = + + =
8 72 4 8 72 4

e. Determine the range of values for changes D1, D2 and D3 respectively in availability for M1, M2 and
M3 for which the optimal dual prices remain valid.

Using the general feasibility conditions obtained above,

1
x1 = 9 + 1/5D1 − 1/8D2 ≥ 0
8

3 5
x2 = 4 − 4/9D1 + D2 ≥ 0
8 8

1 3 1
s3 = 44 + 1 D1 − 2 D2 + D3 ≥ 0
4 8 4

Material M1

For Material M1, availability changes from 100 to 100 + D1. We set other changes to zero i.e. D2 = D3 =
0. Thus,

1 1
x1 = 9 + 1/5D1 − 1/8D2 = 9 + 1/5D1 − 1/8 × 0⟹ D1 ≥ − 45.625
8 8
3 5 3 5
x2 = 4 − 4/9D1 + D2 = 4 − 4/9D1 + × 0 ≥ 0 ⟹ D1 ≤ 9.844
8 8 8 8
1 3 1 1 3 1
s3 = 44 + 1 D1 − 2 D2 + D3 = 44 + 1 D1 − 2 × 0 + 0 ≥ 0 ⟹
4 8 4 4 8 4
D1 ≥ − 60.84

⟹ − 60.84 ≤ D1 ≤ 9.844

-2100 -300 150

Material M2
For Material M2, availability changes from 77 to 77 + D2. We set other changes to zero i.e. D1 = D3 = 0.
Thus,

1 1
x1 = 9 + 1/5D1 − 1/8D2 = 9 + 1/5 × 0 − 1/8D2 ≥ 0 ⟹ D2 ≤ 73
8 8
3 5 3 5
x2 = 4 − 4/9D1 + D2 = 4 − 4/9 × 0 + D2 ≥ 0 ⟹ D2 ≥ − 7 ⟹ − 7 ≤ D2 ≥ 73
8 8 8 8
1 3 1 1 3 1
s3 = 44 + 1 D1 − 2 D2 + D3 = 44 + 1 × 0 − 2 D2 + 0 ≥ 0 ⟹ D2 ≤ 19.67
4 8 4 4 8 4

-140

Material M 3

For Material M3, availability changes from 80 to 80 + D3. We set other changes to zero i.e. D1 = D2 = 0.
Thus,

1 1
x1 = 9 + 1/5D1 − 1/8D2 = 9 + 1/5 × 0 − 1/8 × 0⟹ − ∞ ≤ D3 ≤ ∞
8 8
3 5 3 5
x2 = 4 − 4/9D1 + D2 = = 4 − 4/9 × 0 + × 0 ≥ 0⟹ − ∞ ≤ D3 ≤ ∞ ⟹ D3 ≥ −
8 8 8 8
1 3 1 1 3 1
s3 = 44 + 1 D1 − 2 D2 + D3 = 44 + 1 × 0 − 2 × 0 + D3 ≥ 0 ⟹ D3 ≥ − 44.25
4 8 4 4 8 4
44.25

Summary

Resource Amount (Hours)


Resources Dual Price Feasibility Range
Minimum Current Maximum
Material M1 1 − 60.84 ≤ D1 ≤ 9.844 39.16 100 109.844
Material M2 3/8 − 7 ≤ D2 ≥ 73 70 77 150
Material M3 0 D3 ≥ − 44.25 35.75 80 ∞

Note that the dual prices will remain applicable for any simultaneous changes that keep the solution
feasible, even if the changes violate the individual ranges

e. Suppose there are d1, d2 and d3 in the unit costs (objective coefficients) of products A, B and C
respectively, using the optimal table in (b) determine the reduced costs due to non-basic variables
on the optimal table in terms of the changes.

The Reduced cost per unit is the difference between the consumed cost per unit and the original revenue
per unit:

Reduced Cost = Cost of Consumed resources per unit − Revenue per unit
The optimal tableau reflects the status of the consumed resources as a variable that initially has a higher
value of revenue or cost per unit may not necessarily be in the basis or as strongly used as another as
consumed by the constraint relationships and costs.

Variables
d1 d2 d3 0 0 0
Initial Non-Basic
Initial Basic Variables
Variables
Solution Ratio
Basic x1 x2 x3 s1 s2 s3
Variables

1 0 0 2/3 1 0 122 5/8


z- 3/8
d1 -
1 0 -0 1/5 0 9 1/8
x1 1/8
d2 0 1 4/5 - 4/9 0 4 3/8
x2 5/8
-2
0 0 0 -2 1/8 1 3/8 1 44 ¼
s3 ¼

The Reduced cost for any variable is then computed by summing up the product of the corresponding
entries of the table column of the concerned variable and that the extreme left column entries and the
column entry of the added row on top subtracted from the sum.

Let us do it for the three decision variables here:

d1 d2 d3
Left x1 x2 x3
Column x1 − column × x2 − column × x3 − column ×
x1 − column x2 − column x3 − column
(Left column) (Left column (Left column
2/ 3×1=2
1 0 0 0 0 2/3
/3
d1 1 d1 0 0 0 d1 × 0 = 0
d2 × 4 / 5 = 4
d2 0 0 1 d2 4/5
/ 5d2
1
0 0 0 0 0 -2 1/8 0× −2 =0
8
Subtract d1 d2 d3
0 + d1+0+0 - d1 0+0+d2 + 0 − d2 2 4
Summation + 0 + d2 + 0 − d3
3 5
2 4
Sum 0 0 + d − d3
3 5 2
2 4
+ d − d3
3 5 2

Notice that variables which are basic in the optimal table have no reduced cost, because they do not
deteriorate the optimal value.
Let us look at what happens to the slacks

0 0 0
Left s1 s2 s3
Column s1 − column × s2 − column × s3 − column ×
s1 − column s2 − column s3 − column
(Left column) (Left column (Left column
3

1 3/8
8 0
1 1 1×0= 0
3
=
8
1 1
d1 − × d1 = −
d1 1/5 - 1/8 8 8 0 0 × d1 = 0
5 d1
4 4
d2 × − = − 5
d2 - 4/9 9 9 5/8 5/8 × d2 = d2 0 d2 × 0 = 0
d2 8
11 9 1
0 11/8 ×0=0 -2 ¼ − ×0=0 0×1= 0
8 4
Subtract 0 0 0
d1 4 3 1 5
Summation 1+ + − d2 + 0 − 0 − d + d 0+0+0+0–0
5 9 8 8 1 8 2
d 4 3 1 5
Sum 1 + 1 + − d2 − d1 + d2 0
5 9 8 8 8
Notice again that because s3 is in the optimal basis, its Reduced Cost unit is zero.

f. Determine the optimality range of each of the unit costs (objective coefficients) in the problem for
which the optimal solution remains valid.

Since, we are solving a maximization problem, each of these z-row non-basic coefficients must be greater
or equal to zero (if a minimization problem, less or equal to zero). The optimality conditions
corresponding to x3, s1 and s2 are thus:

2 4
+ d − d3 ≥ 0
3 5 2

d1 4
1+ + − d2 ≥ 0
5 9

3 1 5
− d1 + d2 ≥ 0
8 8 8

These conditions must be satisfied simultaneously to maintain optimality of the current optimum.

g. Determine the ranges of d1, d2 and d3 for which the reduced costs hold.
Treat as in PART (d)

(70 marks)

You might also like