ASSIGNMENT 1

NOTE: Unless specifically stated otherwise, use the method of your choice to obtain a
BFS to the problem. In such cases, it is recommended that you use the Vogel’s
approximation method (VAM), since it is likely to lead to less iterations when you are
trying to find ab Optimal solution:

1. Use the Least (Minimum) Cost method to find a BFS to the following problem and
then determine an Optimal solution where Oi and Dj represent the ith origin and jth
destination respectively

D1 D2 D3 D4 Available D1 D2 D3 D4 A

O1 6 4 1 5 14 O1 14
O2 8 9 2 7 16 O2 16
O3 4 3 6 2 5 O3 5
Requirements 6 10 15 4 R 6 10 1 4
D1 D2 D3 D4 A
O1 14
O2 1 15
O3 5
R 6 10 4

D1 D2 D3 D4 A
O1 14
O2 1 15
O3 1 4
R 6 9

D1 D2 D3 D4 A
O1 14
O2 6 9 1
O3 1 4
R

v1=8 v2=9 v3=2 v4=8

u1=-1 1 4 x 2
u2=0 x x X 1
u3=-6 -2 x -2 x

v1=8 v2=5 v3=2 v4=4

u1=-1 4 1
u2=0 8 2
u3=-2 3 2
 D1 D2 D3 D4 A
v1=8 v2=5 v3=2 v4=4 O1 10 4 14
O2 5 11 16
u1=-1 1 x x -2
O3 1 4 5
u2=0 x -4 x -3
R 6 10 15 4
u3=-2 2 x -6 x

v1=8 v2=5 v3=2 v4=2

u1=-1 1 x x -4
u2=0 x -4 x -5
u3=-4 x -2 -4 x

v1=0 v2=-2 v3=-6 v4=-2

u1=6 6 4
u2=8 8 2
u3=4 4 2

v1=0 v2=-2 v3=-6 v4=-2

u1=6 x x -1 -1
u2=8 x -3 x -1
u3=4 x -1 -8 x

2. A company is to schedule its weekly production of a certain product for the next 4
weeks. The production cost per item is Re. 100. Weekly demands are 400, 800, 800
and 900, which must be met. The company can produce a maximum of 600 units each
week. In addition, the company can employ overtime which increases the production
cost by Re. 20 per item. Excess production can be stored at a unit cost of Re. 10 per
week. How should the production be scheduled so as to minimize the total cost?

The cost matrix will be as follows (see explanation below):

400 800 800 900 5100
P1(R) 100 110 120 130 0 600
P1(O) 120 130 140 150 0 2300
P2(R) M 100 110 120 0 600
P2(O) M 120 130 140 0 1900
P3(R) M M 100 110 0 600
P3(O) M M 120 130 0 1100
P4(R) M M M 100 0 600
P4(O) M M M 120 0 300
The notation Pi(R) means regular production in i th period. Similarly, Pi(O) means
overtime (OT) production in ith period. Production in different periods plays the role
of supply points, and weekly demands similarly play the role of demand points. For
any production that is used to meet demand in a subsequent week, inventory costs are
incurred in addition to production costs, and this is reflected in the cost matrix above.
What is the max that can be produced by OT? It is not specified in the problem; one
must figure it out. We reason as follows: 1. In any period, we must exhaust regular
production before turning to OT (this follows from the cost differential). 2. We cannot
produce (either via regular production or via OT) for a previous week (Note that
backordering should not be assumed unless specifically stated in the problem. In this
case, it is not). 3. For all future periods, we assume the worst: both regular production
and OT (somehow) fail to produce anything.
Our TP now becomes unbalanced, so to balance it, we have to create a dummy
column to absorb extra production (with unit cost 0 for all the cells in the column)

Since backordering is not allowed, we have to ensure that it is excluded. This is done
by assigning a very large (positive) cost (M) to all such allotments. M is assumed to
be big enough that all other cell costs 0 in comparison.

400 800 800 900 5100 We use VAM (one can

P1(R) 100 110 120 130 0 600 100 obviously use any of the three
P1(O) 120 130 140 150 0 2300 120 given methods, per one’s
P2(R) M 100 110 120 0 600 100 choice). All intermediate
P2(O) M 120 130 140 0 1900 120 tableaux are not shown, but the
P3(R) M M 100 110 0 600 100 entire sequence is provided in
P3(O) M M 120 130 0 1100 120 words alongside
P4(R) M M M 100 0 600 100
P4(O) M M M 120 0 300 120
20 10 10 10 0

400 800 800 900 The first three allotments are in

P1(R) 100 110 120 130 0 600 100 cells (2, 5) (4, 5) and (6, 5) of
P1(O) 120 130 140 150 02300 10 the cost matrix (superscripts
P2(R) M 100 110 120 0 600 100 denote cell allotments)
P2(O) M 120 130 140 01900 10
P3(R) M M 100 110 0 600 10
P3(O) M M 120 130 0900 200 10
P4(R) M M M 100 0 600 M-100
P4(O) M M M 120 0 300 M-120
20 10 10 10 0

400 800
P1(R) 100 110 120 130 0 600 10 The next allotments are in
P1(O) 120 130 140 150 02300 10 (7, 4) and (8, 4); (5, 3) and
P2(R) M 100 110 120 0 600 M-100 (6, 3). Ties are resolved
P2(O) M 120 130 140 01900 M-120 arbitrarily. The final
P3(R) M M 100600 110 0 0 allotments are in (3, 2); (1,
P3(O) M M 120200 130 0900 0 1) and (1, 2):
P4(R) M M M 100600 0
P4(O) M M M 120300 0
20 10
 400 200
P1(R) 100400 110200 120 130 0 600 10
P1(O) 120 130 140 150 02300 10
P2(R) M 100600 110 120 0
P2(O) M 120 130 140 01900
P3(R) M M 100600 110 0 0
P3(O) M M 120200 130 0900 0
P4(R) M M M 100600 0
P4(O) M M M 120300 0
20 10

400 800 800 900 5100 This leads to the final allotment shown
P1(R) 400 200 600 on left. Since this is a degenerate
P1(O) 2300 2300 solution, additional cells must be
P2(R) 600 600 allotted a BV.
P2(O) 1900 1900
P3(R) 600 600
P3(O) 200 900 1100
P4(R) 600 600
P4(O) 300 300

400 800 800 900 5100 (Any set of choices that does not
P1(R) 400 200 600 involve a loop is OK, but to cut short
P1(O) 2300 2300 the work, we add an additional
P2(R) 600  600 condition: amongst candidate cells,
P2(O) 1900 1900 choose those that have min cost). Note
P3(R) 600  600 that 8+5-1=12 allotments are
P3(O) 200 900 1100 necessary. (The two symbols  and 
P4(R) 600 600 are used only for differentiation; each
P4(O) 300 300 denotes a basic variable at zero value)

v1=100 v2=110 v3=120 v4=130 v5=0 Next use basic cells to

u1=0 100 110 determine u’s and v’s:
u2=0 0
u3=-10 100 110
u4=0 0
u5=-20 100 110
u6=0 120 0
u7=-30 100
u8=-10 120
For ease cost matrix reproduced below:

v1=100 v2=110 v3=120 v4=130 v5=0

u1=0 x x 120 130 0
u2=0 120 130 140 150 x
u3=-10 M x x 120 0
u4=0 M 120 130 140 x
u5=-20 M M x x 0
u6=0 M M x 130 x
u7=-30 M M M x 0
u8=-10 M M M x 0
 v1=100 v2=110 v3=120 v4=130 v5=0
u1=0 x x 0 0 0 Next test reduced costs of Non-
u2=0 -20 -20 -20 -20 x basic cells to check optimality.
u3=-10 90-M x x 0 -10 The reduced costs are seen to
u4=0 100-M -10 -10 -10 x be as follows, indicating
u5=-20 80-M 90-M x x -20 optimality:
u6=0 100-M 110-M x 0 x
u7=-30 70-M 80-M 90-M x -30
u8=-10 90-M 100-M 110-M x -10
Note: if, instead of In addition, the company can employ overtime which increases the
production cost by Re. 20 per item the problem also had stated This increases the
weekly production by an additional 200 units, this would straightaway have provided
us with the upper limit to the OT production per week (above, we had to deduce this).

3. A department store wishes to purchase the following quantities of ladies’ dresses.

Dress Type A B C D E
Quantity 150 100 75 250 200
Tenders are submitted by 4 different manufacturers who undertake to supply not more
than the quantities below (all types of dresses combined). Required: 775+Dummy 75

Manufacturer W X Y Z
Total Quantity 300 200 150 200
The store estimates that its profit per dress will vary with the manufacturer as shown
in the matrix below. How should orders be placed? Use VAM to get a BFS.

DRESS
A B C D E
W 2.75 3.5 4.25 2.25 1.5
Manufacturer X 3 3.25 4.5 1.75 1
Y 2.5 3.5 4.75 2 1.25
Z 3.25 2.75 4 2.5 1.75
Again, we add a “dummy dress” for which there will be no profit to represent unused
capacity of manufacturers. So the profit matrix is as follows after we initiate VAM (2,
1), (4, 1), (4, 4), (1, 4), (1, 6), (2, 6) and back to (2, 1):

DRESS
Manu A B C D E D’
Fac W 2.75 3.5 4.25 2.25 1.5 0 .75
Turer X 3 3.25 4.5 1.75 1 0 1.25
.
Y 2.5 3.5 4.75 2 1.25 0 1.25
Z 3.25 2.75 4 2.5 1.75 0 .75
 .25 0 .25 .25 .25 0
We follow the VAM procedure through the following iterations. Note that this is a
maximization problem. For BFS, we apply VAM as usual, except that in determining
allocations, we take the difference between max entry and next to max entry, and allot
to the max in the row (or column) where this difference is max.:

150 100 0 250 200 75

DRESS
Manu A B C D E D’
Fac W 2.75 3.5 0 2.25 1.5 0 .75 300
Turer X 3 3.25 0 1.75 1 0 1.25 200
.
Y 2.5 3.5 75 2 1.25 0 1.25 75
Z 3.25 2.75 0 2.5 1.75 0 .5 200
.25 0 .25 .25 0

150 100 0 250 200 75

DRESS
Manu A B C D E D’
Fac W 2.75 3.5 x 2.25 1.5 0 .75 300
Turer X 3 3.25 x 1.75 1 0 .25 200
.
Y 2.5 3.5 75 2 1.25 0 1 75
Z 3.25 2.75 x 2.5 1.75 0 .5 200
.25 0 .25 .25 0

150 250 250 200 75

DRESS
Manu A B C D E D’
Fac W 2.75 3.5 x 2.25 1.5 0 .75 300
Turer X 3 3.25 x 1.75 1 0 .25 200
.
Y 2.5 75 75 2 1.25 0 1 0
Z 3.25 2.75 x 2.5 1.75 0 .5 200
.25 0 .25 .25 0

150 250 250 200 75

DRESS
Manu A B C D E D’
Fac W 2.75 3.5 x 2.25 1.5 0 .75 300
Turer X 3 3.25 x 1.75 1 0 .25 200
.
Y x 75 75 x x x x 0
Z 3.25 2.75 x 2.5 1.75 0 .5 200
.25 .25 .25 .25 0

150 0 0 250 200 75

DRESS
Manu A B C D E D’
Fac W 2.75 25 x 2.25 1.5 0 .5 275
Turer X 3 x x 1.75 1 0 1.25 200
.
Y x 75 75 x x x x 0
Z 3.25 x x 2.5 1.75 0 .75 200
.25 x x .25 .25 0

0 0 0 250 200 75
DRESS
Manu A B C D E D’
Fac W 2.75 25 x 2.25 1.5 0 .5 275
Turer X 150 x x 1.75 1 0 1.25 50
.
Y x 75 75 x x x x 0
Z 3.25 x x 2.5 1.75 0 .75 200
.25 X x .25 .25 0

0 0 0 250 200 75
DRESS
Manu A B C D E D’
Fac W x 25 x 2.25 1.5 0 .75 275
Turer X 150 x x 1.75 1 0 .75 50
.
Y x 75 75 x x x x 0
Z x x x 2.5 1.75 0 .75 200
x x x .25 .25 0

0 0 0 50 200 75
DRESS
Manu A B C D E D’
Fac W x 25 x 2.25 1.5 0 .75 275
Turer X 150 x x 1.75 1 0 .75 50
.
Y x 75 75 x x x x 0
Z x x x 200 x x x 0
x x x .5 .5 0

0 00 0 200 75
DRESS
Manu A B C D E D’
Fac W x 25 x 50 1.5 0 1.5 225
Turer X 150 x x x 1 0 1 50
.
Y x 75 75 x x x x 0
Z x x x 200 x x x 0
x x x x .5 0

0 0 0 0 0 75
 DRESS
Manu A B C D E D’
Fac W x 25 x 50 200 0 25
Turer X 150 x x x x 0 50
.
Y x 75 75 x x x x 0
Z x x x 200 x x x 0
x x x x x 0

So that the final allotment is:

0 0 0 0 0 75
DRESS
Manu A B C D E D’
Fac W x 25 x 50 200 25
Turer X 150 x x x x 50
.
Y x 75 75 x x x
Z x x x 200 x x
0 0 0 0 0 0

Let us now determine optimality. We use the basic cells to determine ui and vj
v
Manu u 3 3.5 4.75 2.25 1.5 0
Fac 0 3.5 2.25 1.5 0
Turer 0 3 0
.
0 3.5 4.75
0.25 2.5

And so for the non-basic cells we have (for costs, consult cost matrix):
DRESS
A B C D E D’
W 0+3-2.75 0+4.75-4.25
X 0+3.5-3.25 0+4.75-4.5 0+2.25-1.75 0+1.5-1
Y 0+3-2.5 0+2.25-2 0+1.5-1.25 0+0-0
Z 0.25+3-3.25 0.25+3.5-2.75 0.25+4.75-4 0.25+1.5-1.75 0.25+0-0

Which upon simplification, yields:

DRESS
A B C D E
W .25 .5
X .25 .25 0.5 0.5
Y 0.5 .25 .25 0
Z 0 1 1 0 0.25
Since this is a max problem, the criterion of optimality is that c ij=ui+vj-cij 0. By this
criterion, we find solution optimal.

4. Powerco has three electric power plants that supply the power needs of four cities.
Plant 1, 2 and 3 can supply the following KW hours of electricity respectively (in
millions): 35, 50 and 40. The peak power demands in these cities, which occur at the
same time, are respectively as follows (in millions): 45, 20, 30, 30. The costs of
sending 1 million KW hours of electricity from plants to cities are as follows (in ’00
rupees):

City 1 City 2 City 3 City 4

Plant 1 8 6 10 9
Plant 2 9 12 13 7
Plant 3 14 9 16 5
(a) Formulate an LP to minimize the cost of meeting each city’s peak power demand
(b) Use the Northwest corner method to obtain a BFS to this problem.
(c) Obtain an optimal solution
Min 8x11+ 6x12+10x13+9x14
+9x21+12x22+13x23+7x24
+14x31 +9x32+16x33+5x34
x11+ x12+x13+x14 =35
x21+ x22+x23+x24 =50
x31+ x32+x33+x34=40
x11 + x21 +x31 =45
x12 + x22 +x32 =20
x13 +x23 +x33 =30
x14 +x24 +x34 =30
all xij0

City 1 City 2 City 3 City 4 A

Plant 1 35 35
Plant 2 10 20 20 50
Plant 3 10 30 40
R 45 20 30 30

v1=8 v2=11 v3=12 v4=1

u1=0 X 5 2 -8
u2=1 X x x -5
u3=4 -2 6 x x
City 1 City 2 City 3 City 4
35
10 10 30
10 30
v1=9 v2=12 v3=13 v4=8
City 1 City 2 City 3 City 4
u1=-1 x 5 2 -2 25 10
u2=0 x x x 1 20 30
u3=-3 -8 x -6 x 10 30
v1=9 v2=7 v3=13 v4=3
u1=-1 x x 2 -7
u2=0 x -5 x -4
u3=2 -3 x -1 x
 v1=9 v2=9 v3=13 v4=5
u1=-3 6 10
u2=0 9 13
u3=0 9 5

v1=9 v2=9 v3=13 v4=5

u1=-3 -2 x x -7 Thereby indicating optimality
u2=0 x -3 x -2
u3=0 -5 x -3 x

5. Consider the following cost and resource matrix. A BFS is also given. Examine
optimality and drive to an optimal solution if not already optimal.
D1 D2 D3 D4 A Clearly this is a
D1 D2 D3 D4
O1 2 3 4 9 20 degenerate solution.
O1 10 10
O2 14 12 5 1 30 Let us therefore
O2  20 10
O3 12 15 9 3 40 include a BV in cell
O3 40
R 10 10 20 50 (2, 2).

v1=11 v2=12 v3=5 V4=1 v1=11 v2=12 v3=5 v4=1

u1=-9 2 3 u1=-9 X x -8 -17
u2=0 12 5 1 u2=0 -3 x x x
u3=2 3 u3=2 1 -1 -2 x

The new solution is optimal. Note that the cell

allotments remain the D1 D2 D3 D4 same (since  denotes zero
allotment) but the O1 10 10 basis changes, since now there is a
different BV at 0 O2 20 10 level (i.e., x13 instead of x22).
O3  40