You are on page 1of 37

chap-5a B.V.

Ramana September 13, 2006 17:59


Chapter 5
Linear Programming
5.1 INTRODUCTION
Optimization problems seek to maximize or min-
imize a function of a number of variables which
are subject to certain constraints. The objective may
be to maximize the prot or to minimize the cost.
The variables may be products, man-hours, money
or even machine hours. Optimal allocation of lim-
ited resources to achieve a given object forms pro-
gramming problems. A programming problem in
which all the relations between the variables is lin-
ear including the function to be optimized is called a
Linear Programming Problem (LPP). G.B. Dantzig,
in 1947, rst developed and applied general problem
of linear programming. Classical examples include
transportation problem, activity-analysis problem,
diet problems and network problem. The simplex
method, developed by GB Dantzig, in 1947, con-
tinues to be the most efcient and popular method to
solve general LPP. Karmarkars method developed
in 1984 has been found to be upto 50 times as fast as
the simplex algorithm. LPP is credited to the works
of Kuhn, Tucker, Koopmans, Kantorovich, Charnes
Cooper, Hitchcock, Stiegler. LPP has been used to
solve problems in banking, education, distribution
of goods, approximation theory, forestry, transporta-
tion and petroleum.
5.2 FORMULATION OF LPP
In a linear programming problem (LPP) we wish to
determine a set of variables known as decision vari-
ables. This is done with the objective of maximiz-
ing or minimizing a linear function of these vari-
ables, known as objective function, subject to certain
linear inequality or equality constraints. These vari-
ables, should also satisfy the nonnegativity restric-
tions since these physical quantities can not be neg-
ative. Here linearity is characterized by proportion-
ality and additivity properties.
Let x
1
, x
2
. . . , x
n
be the n decision unknown vari-
ables and c
1
, c
2
. . . , c
n
be the associated (constant
cost) coefcients. Then the aim of LP is to optimize
(extremise) the linear function,
z = c
1
x
1
+c
2
x
2
+. . . +c
n
x
n
(1)
Here (1) is known as the objective function. (O.F.)
The variables x
j
are subject to the following mlinear
constraints
a
i1
x
1
+a
i2
x
2
+. . . +a
in
x
n

b
i
(2)
for i = 1, 2, ... m. In (2), for each constraint only one
of the signs or or = holds. Finally x
i
should also
satisfy nonnegativity restrictions
x
j
0 for j = 1 to n (3)
Thus a general linear programming problemconsists
of an objective function (1) to be extremized subject
to the constraints (2) satisfying the non-negativity
restrictions (3).
5.1
chap-5a B.V.Ramana September 13, 2006 17:59
5.2 MATHEMATICAL METHODS
Solution
To LPP is any set of values {x
1
, x
2
. . . , x
n
} which
satises all the m constraints (2).
Feasible Solution
To LPP is any solution which would satisfy the non
negativity restrictions given by (3).
Optimal Feasible Solution
To LPP is any feasible solution which optimizes (i.e.
maximizes or minimizes) the objective function (1).
From among the innite number of feasible solu-
tions to an LPP, we should nd the optimal feasible
solution in which the maximum (or minimum) value
of z is nite.
Example 1: Suppose Ajanta clock company pro-
duces two types of clocks standard and deluxe
using three different inputs A, B, C. From the data
givenbelowformulate the LPPtodetermine the num-
ber of standard and deluxe clock to be manufactured
to maximize the prot.
Let x
1
and x
2
be the number of standard and
deluxe clocks to be produced.
Technical coefficients
Input
Standard Deluxe
Capacity
(Resource)
A
B
C
2
2
4
2
4
2
0
3
20
12
16 C
Profit (Rs)
Then the objective function is to maximize the
total prot i.e. maximize z = 2x
1
+3x
2
, since the
prot for one standard clock is Rs 2 and prot for
one deluxe and clock is Rs 3. Because of the limited
resources, for input A we have the following restric-
tion. Since one standard clock consumes 2 units of
resource A, x
1
units of standard clocks consume 2x
1
units of input A. Similarly 4x
2
units of input A is
required to produce x
2
deluxe clocks. Thus the total
requirement of the input Afor production of x
1
, stan-
dard and x
2
deluxe clocks is
2x
1
+4x
2
.
However, the total amount of resource A available is
20 units only. Therefore the restriction on resource
A is
2x
1
+4x
2
20
Similarly the restriction of resource B is
2x
1
+2x
2
12
and restriction on source C is
4x
1
16.
Since x
1
, x
2
are physical quantities (the number of
clocks produced), they must be non negative i.e.
x
1
0
and x
2
0.
Thus the LPP consists of the O.F., three inequality
constraints and the non-negativity restrictions.
5.3 GRAPHICAL SOLUTION OF LPP
When the number of decision variables (or products)
is two, the solution to linear programming problem
involving any number of constraints can be obtained
graphically. Consider the rst quadrant of the x
1
x
2
plane since the two variables x
1
and x
2
should sat-
isfy the nonnegativity restrictions x
1
0 and x
2
0.
Now the basic feasible solution space is obtained in
the rst quadrant by plotting all the given constraints
as follows. For a given inequality, the equation with
equality sign (replacing the inequality) represents a
straight line in x
1
x
2
plane dividing it into two open
half spaces. By a test reference point, the correct side
of the inequality is identied. Say choosing origin
(0, 0) as a reference point, if the inequality is satis-
ed then the correct side of the inequality is the side
on which the origin (0, 0) lies. Indicate this by an
arrow. When all the inequalities are plotted like this,
in general, we get a bounded (or unbounded in case
of greater than inequalities) polygon which enclose
the feasible solution space, any point of which is a
feasible solution.
For z = z
0
, the objective functionz = c
1
x
1
+c
2
x
2
represents an iso-contribution (or iso-prot) straight
line say x
2
=
c
1
c
2
x
1
+
z
0
c
2
_
or x
1
=
c
2
c
1
x
2
+
z
c
1
_
chap-5a B.V.Ramana September 13, 2006 17:59
LINEAR PROGRAMMING 5.3
such that for any point on this line, the contribu-
tion (Prot) (value of z
0
) is same. To determine the
optimal solution, in the maximization case, assign-
ing arbitrary values to z, move the iso-contribution
line in the increasing direction of z without leav-
ing the feasible region. The optimumsolution occurs
at a corner (extreme) point of the feasible region.
So the iso-prot line attains its maximum value of
z and passes through this corner point. If the iso-
contribution line (objective function) coincides with
one of the edges of the polygon, then any point on
this edge gives optimal solution with the same max-
imum (unique) value of the objective function. Such
a case is known as multiple (alternative) optima case.
In the minimization case, assigning arbitrary values
to z, move the iso-contribution line in the direction
of decreasing z until it passes through a corner point
(or coincides with an edge of the polygon) in which
case the minimum is attained at this corner point.
Range of Optimality
For a givenobjective functionz = c
1
x
1
+c
2
x
2
, slope
of z changes as the coefcients c
1
and c
2
change
which may result in the change of the optimal corner
point itself. In order to keep (maintain) the current
optimum solution valid, we can determine the range
of optimality for the ratio
c
1
c
2
_
or
c
2
c
1
_
by restricting the
variations for both c
1
and c
2
.
Special Cases:
(a) The feasible region is unbounded and in the case
of maximization, has an unbounded solution or
bounded solution.
(b) Feasible region reduces to a single point which
itself is the optimal solution. Such a trivial solu-
tion is of no interest since this can be neither max-
imized nor minimized.
(c) A feasible region satisfying all the constraints is
not possible since the constraints are inconsistent.
(d) LPP is ill-posed if the non-negativity restriction
are not satised although all the remaining con-
straints are satised.
Examples:
(a) x
1
+3x
2
3, x
1
+x
2
2, x
1
, x
2
0,
Maximize: z = 1.5x
1
+2.5x
2
unbounded feasi-
ble region, unbounded solution (can be maxi-
mized indenitely)
Fig. 5.1
(b) x
1
x
2
0, 0.5x
1
+x
2
1 x
1
, x
2
0
Maximize: z = x
2
0.75x
1
unbounded feasible
region z
2
= 0.5 is bounded optimal solution.
Fig. 5.2
(c) x
1
+x
2
2, x
1
, 5x
2
10
Maximize: z = 5x
2
, x
1
, x
2
0, (0, 2) is unique
solution, max: z = 10.
(0, 2)
Fig. 5.3
(d) x
1
+x
2
1, 0.5x
1
5x
2
10
Maximize: z = 5x
2
, x
1
, x
2
0. No feasible
region. Constraints are inconsistent
chap-5a B.V.Ramana September 13, 2006 17:59
5.4 MATHEMATICAL METHODS
Fig. 5.4
(e) 1.5x
1
+1.5x
2
9, x
1
+x
2
2.
No feasible region.
(0, 6)
(6, 0)
(0, 2)
(2, 0)
WORKED OUT EXAMPLES
Example 1: ABC company produces two types of
calculators. A business calculator requires 1 hour of
wiring, one hour of testing and 3 hours of assem-
bly, while a scientic calculator requires 4 hours of
wiring, one hour of testing and one hour of assembly.
A total of 24 hours of wiring, 21 hours of assembly
and 9 hours of testing are available with the company.
If the company makes a prot of Rs 4 on business
calculator (BC) and Rs 10 on scientic calculator
(SC), determine the best product mix to maximize
the prot.
Solution: Let x
1
be the number of business cal-
culators (BC) produced while x
2
be the number of
scientic calculators (SC) produced. Then the objec-
tive is to maximize the prot z = 4x
1
+10x
2
subject
to the following ne constraints:
x
1
+4x
2
24 (wiring) (I)
x
1
+x
2
9 (testing) (II)
3x
1
+x
2
21 (assembly) (III)
x
1
0
x
2
0
_
non negative (IV)
constraints (V)
To determine the feasible solution space consider
the rst quadrant of the x
1
x
2
-plane since x
1
0
and x
2
0. Then draw the straight lines x
1
+4x
2
=
24, x
1
+x
2
= 9 and 3x
1
+x
2
= 2.1. Note that an
inequality divides the x
1
x
2
-plane into two open half-
space. Choose any reference point in the rst quad-
rant. If this reference point satises the inequality
then the correct side of the inequality is the side on
which the reference point lies. Generally origin (0,
0) is taken as the reference point. The correct side of
the inequality is indicated by an arrow. The shaded
region is the required feasible solution space satisfy-
ing all the ve constraints. The ve corner points of
the feasible region are A(0, 0), B(7, 0), C(6, 3), D(4,
5), E(0, 6). Identify the direction in which z increases
without leaving the region. Arbitrarily choosing z =
0, 28, 54, 60, 66, observe that the straight lines (prot
function) z = 4x
1
+10x
2
or x
2
=
2
5
x
1
+
z
10
passes
through the corner points A, B, C, E, D respectively.
The optimumsolution occurs at the corner point D(4,
5), where the maximum value for z = 66 is attained.
Thus the best product mix is to produce 4 business
and 5 scientic calculator which gives a maximum
prot of Rs. 66.
x
2
Solution
Space
3
+
=
2
1
x
x
1
2
x
x
1
2
+
4
=
2
4
x
x
1
2
+
=
9
3
1
2
Range of
optimality
I
n
c
r
e
a
s
i
n
g
z
z
=
0
z
=
2
8
z
=
5
4
E
z
=
6
0
z
=
6
6
A B
C
D
x
1
O
p
t
i
m
u
m
Fig. 5.5
Example 2: (a) Solve the above problems to min-
imize z = 4x
1
10x
2
.
(b) If z = c
1
x
1
+c
2
x
2
, does an alternative optimal
solution exists
(c) Determine the range of optimality for the ratio
c
1
c
2
_
or
c
2
c
1
_
.
chap-5a B.V.Ramana September 13, 2006 17:59
LINEAR PROGRAMMING 5.5
Solution: (a) Rewriting, x
2
=
2
5
x
1

z
10
. Choose
z = 0, 28, 54, 60, 66, then the objective
function passes through the corner points A, B, C,
E, Drespectively. Thus the minimumvalue z = 66
is attained at the corner point D(4, 5). Observe that
maximum of z = 4x
1
+10x
2
is 66 and minimum of
z = 4x
1
10x
2
is 66
i.e., max f (x) = min (f (x)).
(b) If z = c
1
x
1
+c
2
x
2
coincides with the straight line
CD: x
1
+x
2
= 9, then any point on the line segment
CD is an optimal solution to the current problem and
thus has multiple (innite) alternative optima.
(c) Let z = c
1
x
1
+c
2
x
2
be the objective function.
Then for c
2
= 0, we write this as
x
2
=
c
1
c
2
x
1
+
z
c
2
The straight line
x
1
+4x
2
= 24 rewritten as x
2
=
1
4
x
1
+
24
4
has
slope
1
4
and the straight line x
1
+x
2
= 9 rewritten
as x
2
= x
1
+9 has slope 1. Thus range of opti-
mality which will keep the present optimumsolution
valid is
1
4

c
1
c
2
1
For c
2
= 4, 1 c
1
4
Similarly for c
1
= 0, the range of optimality is
1
c
2
c
1
4. For c
1
= 2, 2 c
2
8.
Example 3: The minimumfertilizer needed/hector
is 120 kgs nitrogen, 100 kgs phosphorous and 80 kgs
of potassium. Two brands of fertilizers available have
the following composition.
Fertilizer Nitrogen Phos. Potassium Price/100 kgs bag
A 20% 10% 10% Rs 50
B 10% 20% 10% Rs 40
Determine the number of bags of fertilizer A and
B which will meet the minimum requirements such
that the total cost is minimum.
Solution: Let X be the number of bags of fertilizer
Apurchased and Y be the number of bags of fertilizer
B purchased. Then the objective is to minimize the
total cos t = z = 50X +40Y
subject to
20X +10Y 120 (Nitrogen)
10X +20Y 100 (Phosphorous)
10S +10Y 80 (Potassium)
and X, Y 0
Draw the straight lines
2X +Y = 12 (1)
X +2Y = 10 (2)
X +Y = 8 (3)
A(0, 12), B(4, 4), C(6, 2), D(10, 0)
X
2
2
3
3
1
1
z
=
5
0
0
z
=
4
8
0
z
=
3
8
0
z
=
3
6
0
C
B
A
Y
Decreasing
direction
of z
(Unbounded)
Feasible
region
D
Fig. 5.6
Objective function: iso-prot equation:
Y =
5
4
X +
z
40
(4)
Choose z = 500, 480, 380, 360 then (4) passes
through the corner points D, A, C, B respectively.
Thus the optimal solution occurs at B(4, 4) i.e. pur-
chase four bags of fertilizer Aand 4 bags of fertilizer
B with a total minimum cost of Rs 360/-.
EXERCISE
Solve the following LPP graphically:
1. Right Wood Furniture Company manufactures
chairs and desks. The time required (in minutes)
and the total available time is given below. If com-
pany sells a chair for a prot of Rs. 25 and desk
chap-5a B.V.Ramana September 13, 2006 17:59
5.6 MATHEMATICAL METHODS
for a prot of Rs 75/- determine the best product
mix that will maximize the prot.
Chair Desk
Available
time
Fabrication
Assembly
Upholstery
Linoleum
15
12
18.75

40
50

56.25
27,000
27,000
27,000
27,000
Ans: Produce 1000 chairs and 300 desks, making a
prot of Rs 47,500.
Hint: Corner points are A(0, 0), B(1440, 0),
C(1440, 135), D(1000, 300), E(250, 480), F(0,
480): Maximize: z = 25X + 75Y. s.t.
15X +40Y 27, 000,
12X +50Y 27000,
18.75X 27000, 56.25Y 27000
2. Asia paints produces two types of paints with the
following requirements.
Standard
Paint
Delux
Paint
Total Available
Quantity (in tons)
Base
Chemicals
Profit (in 100s)
6
1
5
4
2
4
24
6
Determine the optimum (best) product mix of the
paints that maximizes the total prot for the com-
pany. Demand for deluxe paint can not exceed
that of standard paint by more than 1 ton. Also
maximum demand of deluxe paint is 2 tons.
Ans: Produce 3 tons of standard and 1.5 tons of deluxe
paint, making a prot of Rs 2100.
Hint: Corner points: A(0, 0), B(4, 0), C(3, 1.5),
D(2, 2), E(1, 2), F(0, 1); OF: Maximize z = 5x
1
+
4x
2
subject to
6x
1
+4x
2
24, x
1
+2x
2
6,
x
1
+x
2
1, x
2
2, x
1
, x
2
0.
3. In an oil renery, two possible blending processes
for which the inputs and outputs per production
run are given below.
Process
I
II
Input
Crude Crude
5 3
4 5
A B
Output
Gasoline Gasoline
5 8
4 4
X Y
Amaximum of 200 units of crude Aand 150
units of crude B are available. It is required to
produce at least 100 units of gasoline X and 80
units of gasoline Y. The prot fromprocess I is Rs
300 while from process II is Rs 400. Determine
the optimal mix of the two processes.
Ans: Produce 30.7 units by process I and 11.5 units
from process II, getting a maximum prot of Rs
13,846.20.
Hint: Maximize: z = 300x
1
+400x
2
, subject to
5x
1
+4x
2
200, 3x
1
+5x
2
150
5x
1
+4x
2
100, 8x
1
+4x
2
80
Corner points: A(20, 0), B(40, 0), D(0, 30),
E(0, 25), C
_
400
13
,
150
13
_
4. Minimize z = 0.3x
1
+0.9x
2
subject to
x
1
+x
2
800, 0.21x
1
0.30 x
2
0, 0.03x
1

0.01 x
2
0, x
1
, x
2
0
Ans: x
1
= 470.6, x
2
= 329.4, minimum cost: Rs
437.64.
5. Maximize: z = 30x
1
+20x
2
subject to x
1
60,
x
2
75, 10x
1
+8x
2
800.
Ans: x
1
= 60, x
2
= 25, Max: prot = Rs 2300
Hint: Corner points: A(0, 0), B(60, 0), C(60, 25),
D(20, 75), E(0, 75).
6. Given x
1
0, x
2
0, x
1
+2x
2
8, 2x
1
x
2

2 solve to (a) max x
1
(b) max x
2
(c) min x
1
(d)
min x
2
(e) max 3x
1
+2x
2
(f) min 3x
1
2x
2
(g)
max 2x
1
2x
2
Ans: (a) x
1
= 8 (b) x
2
=
18
5
(c) x
1
= 0 (d) x
2
= 0 (e) z
= 24, x
1
= 8, x
2
= 0 (f) z = 24, x
1
= 8, x
2
= 0
(g) z =
28
5
, x
1
=
4
5
, x
1
=
18
5
7. Minimize z = x
1
+x
2
s.t. x
1
0,
x
2
0 2x
1
= x
2
12, 5x
1
+8x
2
74,
x
1
+6x
2
28.
Ans: x
1
= 2, x
3
= 8, min. 10
Hint: Unbounded region with corner points
A(0, 12), B(2, 8), C(10, 3), D(28, 0)
chap-5a B.V.Ramana September 13, 2006 17:59
LINEAR PROGRAMMING 5.7
8. Maximize: z = 5x
1
+3x
2
s.t. x
1
0, x
2
0,
3x
1
+5x
2
15, 5x
1
+2x
2
10
Ans: x
1
= 1.053, x
2
= 2.368, Max: 12.37
Hint: Corner points: (0, 3), (1.053, 2.368), (2, 0)
9. Maximize: z = 2x
1
4x
2
s.t. x
1
0, x
2
0,
3x
1
+5x
2
15, 4x
1
+9x
2
36
Ans: x
1
= 9, x
2
= 0, Max: 18
Hint: Corner point: (0, 3), (0, 4), (5, 0) (9, 0)
10. Maximize: z = 3x
1
+4x
2
s.t. x
1
0, x
2
0,
2x
1
+x
2
40, 2x
1
+5x
2
180
Ans: x
1
= 2.5, x
2
= 35, z = 147.5
Hint: Corner points: 0(0, 0), A(20, 0), B(2.5, 35),
C(0, 36)
11. Minimize z = 6000x
1
+4000x
2
s.t. x
1
0, x
2

0, 3x
1
+x
2
40, x
1
+2.5x
2
22, x
1
+x
2

40
3
.
Ans: x
1
= 12, x
2
= 4, z
min
= 88,000
Hint: A(22, 0), B(12, 4), C(0, 40)
Note: Constraint x
1
+x
2

40
3
is redundant.
12. Maximize z = 45x
1
+80x
2
s.t. 5x
1
+20x
2

400, 10x
1
+15x
2
450.
Ans: x
1
= 24, x
2
= 14, z = Rs2200.
5.4 CANONICAL AND STANDARD FORMS
OF LPP
Since max f (x) = min (f (x)), an LPP with
maximation can be transferred to a minimization
problemand vice versa. Thus, the following analysis
can be applied for a maximization or minimization
problem without any loss of generality.
Canonical form of LPP is an LPP given by (1) (2)
(3) with all the constraints (2) are of the less than or
equal to type.
Standard form of LPP consists of (1) (2) (3) with all
constraints (2) are of the equality type and with all
bi 0, for i = 1 to m.
Conversion to Standard Form Given any general
LPP, it can be transformed to standard LPP as fol-
lows:
1. In any constraint if the right hand side con-
stant b
i
is negative, then multiply that constraint
throughout by 1. (Note that multiplication of an
inequality constraint by 1, reverses, the inequal-
ity sign i.e. 3 < 2, multiplied by 1 we get
(1)(3) > (1)(2) or 3 > 2.
2. A less than or equal to type constraint

j
a
ij
x
j
b
i
; (b
i
0) gets transformed to an
equality

j
a
ij
x
j
+s
i
= b
i
by the addition of a slack variable s
i
, which
is non negative.
3. A greater than or equal to type constraint

j
a
ij
x
j
b
i
; (b
i
0)
can be transformed to an equality

a
ij
x
j
s
i
= b
i
by subtracting a surplus variable S
i
, which
is non negative. In general, it is more convenient
to work with equations rather than with inequal-
ities. So given any general LPP, convert it to a
standard LPP, consisting of m simultaneous lin-
ear equations in "n" unknown decision variables.
Minimize:
z = c
1
x
1
+c
2
x
2
+ +c
n
x
n
(1)
subject to
a
11
x
1
+a
12
x
2
+ +a
1n
x
n
= b
1
a
21
x
1
+a
22
x
2
+ +a
2n
x
n
= b
2

a
m1
x
1
+a
m2
x
2
+ +a
mn
x
n
= b
m

(2)
and x
1
, x
2
, x
3
, x
n
0 (3)
Here c
j
(Prices), b
j
(requirements) and a
i,j
(activity coefcients) for i = 1 to m, j = l to n)
are known constants.
If m > n, discard the mn redundant equa-
tions. If m = n, the problem may have a unique
(single) solution which is of no interest since
it can neither be maximized or minimized. If
m < n, which ensures that none of the equations
is redundant, then there may exist innite number
of solutions from which an optimal solution can
be obtained.
Assume that m < n. Set arbitrarily any
n m variables equal to zero and solve the m
equations for the remaining m unknowns. Sup-
pose the unique solution obtained be
{x
1
, x
2
, x
m
}, by setting the remaining
(n m) variables
x
m+1
, , x
n
all to zero.
chap-5a B.V.Ramana September 13, 2006 17:59
5.8 MATHEMATICAL METHODS
Basic solution
{x
1
, x
2
, , x
m
} is the solution of the systemof equa-
tions (2) in which n m variables are set to zero.
Basic variables
are the variables x
1
, x
2
, , x
m
in the basic solution.
Basis
is the set of m basic variables in the basic solution.
Non-basic variables
x
m+1
, x
m+2
, x
n
are the (n m) variables which
are equated to zero to solve the m equations (2),
(resulting in the basic solution).
Basic feasible solution
is a basic solution which satises the nonnegativity
restrictions, (3) i.e. all basic variables are non nega-
tive. (i.e. x
j
0 for j = 1, 2, 3, m)
Nondegenerate basic feasible solution
is a basic feasible solution in which all the basic vari-
ables are positive (i.e., x
j
> 0 for j = 1, 2, 3, m)
Optimal basic feasible solution
is a basic feasible solution which optimizes (in this
case minimizes) the objective function (1).
Why Simplex Method
In an LPP with m equality constraints and n vari-
ables with m < n, the number of basic solutions is
n
c
m. For small n and m, all the basic solutions (cor-
ner points) can be enumerated (listed out) and the
optimal basic feasible solution can be determined.
Example:
Maximize: z = 2x
1
+3x
2
s.t. 2x
1
+x
2
4, x
1
+
2x
2
5. Rewriting 2x
1
+x
2
+x
3
= 4, x
1
+2x
2
+
x
4
= 5. Here m = 2, n = 4, n
c
m = 4
c
2 = 6 The six
basic solutions are: 1. (0, 0, 4, 5), Feasible (F), Non-
degenerate (ND) and z = value of O.F = 0
2. (0, 4, 0, -3), NF (non feasible)
3. (0, 2.5, 1.5, 0), z = 7.5 F, ND
4. (2, 0, 0, 3), z = 4, F, ND
5. (5, 0, - 6, 0), NF
6. (1, 2, 0, 0), z = 8,
Feasible nondegenerate and optimal.
However, even for n = 20, m = 10, the number
of basic solutions to be investigated is 1,84,756, a
large part of which are infeasible. It is proved that
the set of feasible solutions to a LPP form a convex
set (the line joining any two points of the set lies in
the set) and the corner (extreme) points of the convex
set are basic feasible solutions. If there is an optimal
solution, it exists at one of these corner points. The
simplex method devised by GB Dantzig is a power-
ful procedure which investigates in a systematic way
for optimal solution at these corner points which are
nite in number.
For m = 10, n = 20, simplex method obtains the
optimal in 15 steps, thus having an advantage of
92,378 to 1.
5.5 SIMPLEX METHOD
The simplex method is an algebraic iterative pro-
cedure which solves any LPP exactly (not approxi-
mately) or gives an indication of an unbounded solu-
tion. Starting at an initial extreme point, it moves
in a nite number of steps, between m and 2m, from
one extreme point to the optimal extreme point. Con-
sider the following LPP with m less than or equal
to inequalities in n variables.
Maximize z = c
1
x
1
+c
2
x
2
+ +c
n
x
n
subject to a
11
x
1
+a
12
x
2
+ +a
1n
x
n
b
1
a
21
x
1
+a
22
x
2
+ +a
2n
x
m
b
2
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
a
m1
x
m1
+a
m2
x
2
+ +a
mn
x
n
b
m
Introducing m slack variables s
1
, s
2
, . . ., s
m
, the
less than or equal to in equalities are converted to
equations.
a
11
x
1
+a
12
x
2
+ +a
1n
x
n
+s
1
= b
1
a
21
x
1
+a
22
x
2
+ +a
2n
x
n
+s
2
= b
2
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
a
m1
x
1
+a
m2
x
2
+ +a
mn
x
n
+ +s
m
= b
m
Here x
1
, x
2
, , x
n
, s
1
, s
2
, , s
m
are all nonneg-
ative i.e. 0. The objective function is rewritten as
chap-5a B.V.Ramana September 13, 2006 17:59
LINEAR PROGRAMMING 5.9
Maximize: z = c
1
x
1
+ +c
n
x
n
+0.s
1
+ +
0.s
m
.
Thus there are m equations in m+n variables.
Putting (m+n) m = n variables zero we get a
starting basic feasible solution. Take x
1
= x
2
=
x
n
= 0. Then the initial solution contains the m
basic variables s
1
, s
2
, s
m
. This corresponds to the
corner point origin with value of the objective func-
tion zero. Since this is a problem of maximization,
the value of objective function will increase if we
introduce one of non-basic variable x
j
(j = 1 to n),
into the solution forcing out one of the basic variable.
The obvious choice is the x
j
with the largest c
j
. Ties
are broken arbitrarily. The objective equation is writ-
ten as z c
1
x
1
c
2
x
2
c
n
x
n
+0.s
1
+ +
0.s
n
= 0
For efcient use, this data is written in the form of
a table known as simplex tableau shown below:
Remark Basis Solution
-row 1 0 0 0 0
-row 0 1 0 0
s -row 0 0 1
-row
-row 0 0 0 1
z x x x x
n
s s s
c z c c c c
s s a a a a b
s a a a a b
s a
s s a a a a b
1 2 1 2
1 2
1 1 11 12 1 1 1
2 2 21 22 2 2 2
1 2
j m
j n
j n
j n
i ij
m m m m mj mn m
1

The rst row, z-row contains the coefcients of


the objective equation with last element in rectangle
indicating the current value of the objective func-
tion (In the present case it is zero). The left most
(rst) column indicates the current basic variables
s
1
, s
2
, , s
m
. The right most (last) column is the
solution column. Thus s
1
= b
1
, s
2
= b
2
, , s
m
=
b
m
(all resources unused) is the basic solution with
the value of the objective function zero.
Test for optimality
If all the z-row coefcients of the nonbasic variables
are nonnegative, then the current solution is optimal.
Stop. Otherwise goto step I.
Step I. Entering variables: Suppose - c
j
, the z-row
coefcient of the non basic variables x
j
is the most
negative, then the variable x
j
will enter the basis. The
jth column is known as the pivotal column.
Step II. Leaving variable: Divide the solution col-
umn with the corresponding elements of the pivotal
column, with strictly positive denominator. Ignore
the ratios, when the pivotal column elements are zero
or negative.
Suppose
b
i
a
ij
is the smallest non negative ratio
among these ratios
b
i
a
ij
,
b
2
a
2j
,
b
m
a
mj
,
then the basic variable s
i
will leave the basis (and
therefore will become a non basic variable). The ith
row is known as the pivotal row. The element a
ij
at the intersection of the pivotal column and pivotal
row is known as the pivotal element, which is encir-
cled in the table step III. Compute the new simplex
tableau with (s
1
, s
2
, , x
j
, , s
m
) as the newbasis
compute.
Pivot row:
New pivot row =
current pivot row
pivot element
.
All other rows including z:
New row = current row - (Corresponding pivot
column coefcient) (New pivot row).
The solution-column in the new tableau readily
gives the newbasic solution with newobjective value
(last element in the z-row). Now test for optimality.
If yes, stop. Otherwise go to step I.
Optimality condition
The nonbasic variable having the most negative (pos-
itive) coefcient coefcient in the z-row will be the
entering variable in a maximization (minimization)
problem. Ties are broken arbitrarily. When all the z-
row coefcients of the non basic variables are non-
negative (nonpositive) then the current solution is
optimal.
Feasibility condition
In both the maximization and minimization prob-
lems, the basic variable associated with the smallest
nonnegative ratio (with strictly positive denomina-
tor) will be the leaving variable.
chap-5a B.V.Ramana September 13, 2006 17:59
5.10 MATHEMATICAL METHODS
Thus the simplex method can be summarized as
follows:
Step 0. If all the constraints are less than or equal
to type, introduce slack variables and determine the
starting basic solution.
Step I. Using optimality condition, select the enter-
ing variable. If no variable can enter the basis, stop.
The current solution is optimal.
Step II. Using feasibility condition determine the
leaving variable.
Step III. Compute the new basic solution (new sim-
plex tableau) and go to step I.
Articial Variable Technique
For a LPP in which all the constraints are less than or
equal to type with b
i
0, an all-slack, initial basic
feasible solution readily exists. However for prob-
lems involving inequalities or equality constrains
no such solution is possible. To alleviate this, arti-
cial variables are introduced in each of the or =
type constraints, and slack variables for the less than
or equal to type which will then provide a starting
solution. The M-method and the two-phase method
are two closely related methods involving articial
variables.
M-Method (also Known as Charnes Method
or Big M-Method)
Since articial variables are undesirable, the coef-
cient for the articial variable in the objective func-
tion is taken as M in maximization problem and
as + M in minimization problems. Here M is a very
large positive (penalty) value. The augmented prob-
lem is solved by simplex method, resulting in one of
the following cases:
1. When all the articial variables have left the basis
and optimality condition is satised, then the cur-
rent solution is optimal.
2. When one or more articial variables are present
in the basis at zero level and the optimality condi-
tion is satised, then the solution is optimal with
some redundant constraints
3. No feasible solution exists when one or more arti-
cial variables are present in the basis at a positive
level although the optimality condition is satis-
ed. Such a solution is known as pseudo optimal
solution since it satises the constraints but does
not optimize the objective function.
Note: Since articial variables which is forced out
of the basis, is never considered for reentry, the col-
umn corresponding to the articial variable may be
omitted from the next simplex tableau.
Two-Phase Method
In the M-method, M must be assigned some specic
numerical value which creates trouble of roundoff
errors especially in computer calculations. The z-
coefcient of the articial variable will be of the form
aM + b. For large chosen M, b may be lost and for
small chosen Mand small a, b may be present leading
to incorrect results. The two phase method consists
of two phases and alleviates the difculty in the M-
method.
Phase I
Exactly as in M-method, introduce necessary arti-
cial variables to get an initial basic feasible solution.
Solve this augmented problem, by simplex method
to minimize r, the sum of the articial variables. If
r = 0, then all the articial variables are forced out
of the basis. Goto phase II. If r > 0, indicating the
presence of articial variables at non zero level, LP
has no feasible solution
Phase II
The feasible solution of phase I forms the initial basic
feasible solution to the original problem(without any
articial variables). Apply simplex method to obtain
the optimal solution.
WORKED OUT EXAMPLES
Enumeration
Example 1: Solve the following LPP by enumerat-
ing all basic feasible solutions. Identify the infeasible
chap-5a B.V.Ramana September 13, 2006 17:59
LINEAR PROGRAMMING 5.11
solutions. Find the optimal solution and the value of
the objective function.
Maximize z = 2x
1
+3x
2
+4x
3
+7x
4
subject to
2x
1
+3x
2
x
3
+4x
4
= 8
x
1
2x
2
+6x
3
7x
4
= 3
and x
1
, x
2
, x
3
, x
4
0.
Solution: The number of equations m = 2. The
number of variable n = 4. The number of basic vari-
ables = m = 2. The number of all possible solutions
is
4
c
2
= 6.
1. Put x
3
= x
4
= 0, solving 2x
1
+3x
2
= 8, x
1

2x
2
= 3, we get x
1
= 1, x
2
= 2, z = 8. Basic
feasible solution, not optimal, x
1
, x
2
are basic
variables, x
3
, x
2
are non basic variables (which
are always zero).
2. Put x
2
= x
4
= 0. Solving 2x
1
x
3
= 8, x
1
+
6x
3
= 3, we get x
1
=
45
13
, x
3
=
14
3
. Since x
3
<
0, this is a basic non feasible solution.
3. Put x
1
= x
4
= 0. Solving 3x
2
x
3
= 8, 2x
2
+
6x
3
= 3, we get x
2
=
45
16
, x
3
=
7
16
, z =
163
16
. This
is a basic feasible solution (not optimal).
4. Put x
3
= x
2
= 0, solving 2x
1
+4x
4
= 8, x
1

7x
4
= 3, we get x
1
=
22
9
, x
4
=
7
9
, z =
93
9
, basic
feasible solution (not optimal).
5. Put x
1
= x
3
= 0. Solving 3x
2
+4x
4
= 8, 2x
2
+
7x
4
= 3 we get x
2
=
44
13
, x
4
=
7
13
. This is a basic
non feasible solution.
6. Put x
1
= x
2
= 0. Solving x
3
+4x
4
= 8, 6x
3

7x
4
= 3, we get x
3
=
44
17
, x
4
=
45
17
. Thus the opti-
mal basic feasible solution with the basic vari-
ables x
3
=
44
17
, x
4
=
45
17
(and obviously the remain-
ing non basic variables x
1
, x
2
at zero value) has
the maximum value of the objective function as
491
17
.
Simplex Method: Maximization
Example 1: Solve the following LPP by simplex
method.
Maximize z = 2x
1
+3x
2
subject to 2x
1
+4x
2
20
2x
1
+2x
2
12
4x
1
16
x
1
0, x
2
0
Solution: Introducing three slack variables, the
giventhree less thanor equal toinequalityconstraints
will be expressed as equations. Assign zero cost to
each of these slack variables. Then the standard form
of the LPP is to
Maximize z = 2x
1
+3x
2
+0 s
1
+0 s
2
+0 s
3
subject to
2x
1
+4x
2
+s
1
=20
2x
1
+2x
2
+s
2
=12
4x
1
+s
3
=16
and x
1
, x
2
, s
1
, s
2
, s
3
0
Express the objective equation as
z 2x
1
3x
2
= 0
Then the starting simplex tableau is represented as
follows:
Basis Solution Remark
1 2 3 0 0 0 0 -row
0 2 4 1 0 0 20 -row
0 2 2 0 1 0 12 -row
0 4 0 0 0 1 16 -row
z x x s s s
z z
s s
s s
s s
1 2 1 2 3
1 1
2 2
3 3
Corner points: A(0, 0), B(4, 0), C(4, 2), D(2, 4), E(0,
5). Value of O.F. at these extreme points: z
A
= 0,
z
B
= 8, z
C
= 14, z
D
= 16, z
E
= 15
2
x
1
1
3
B A
E
x
2
D
optimum ( = 2, = 4)
= 16
x x
z
1 2
D
C
Fig. 5.2
The initial basis consists of the three basic vari-
ables s
1
= 20, s
2
= 12, s
3
= 16. The two non basic
variables are x
1
= 0, x
2
= 0. Note that non basic vari-
ables are always equal to zero. Thus this solution
corresponds to the corner (extreme) point A (0, 0) in
the graph. In the simplex tableau all the three basic
chap-5a B.V.Ramana September 13, 2006 17:59
5.12 MATHEMATICAL METHODS
variables are listed in the left-most (rst) column
and their values (including the value of the objective
function), in the right-most (last) column. Here the
value of OF is 0 since all the resources are unutilized.
In the z-row, the value of the objective function in the
solution column is enclosed in a square. Since this is
a maximization problem, to improve (increase) the
value of z, one of the non-basic variables will enter
into the basis and there by forcing out one of the
current basic variable from the basis (since the num-
ber of basic variables in the basis is xed and equals
to m = 3 the number of constraints). From the opti-
mality condition, the entering variable is one with
the most negative coefcient in the z-row. In the z-
row the most negative elements is 3. Thus the non
basic variable x
2
will enter the basis. Todetermine the
leaving variable, calculate the ratios of the right-hand
side of the equations (solution-column) to the corre-
sponding constraint coefcients under the entering
variable x
2
, as follows:
Basis Entering
x
2
Solution Ratio
20
4
12
2
16
0
= 5 minimum
= 6
= (Ignore)
S
1
S
2
S
3
4
2
0
20
12
16
Therefore s
1
is the leaving variable. The value of
the entering variable x
2
in the new solution equals to
this minimum ratio 5. Here s
1
-row is the pivot row;
x
2
column is the pivot column and the intersection
of pivot column and pivot row is the pivot element
4 which is circled in the tableau. The new pivot row
is obtained by dividing the current pivot row by the
pivot element 4. Thus the new pivot row is
0
4
2
4
4
4
1
4
0
4
0
4
20
4
i.e., 0
1
2
1
1
4
0 0 5
Recall that for all other rows, including z-row,
New row = current row (corresponding pivot
coefcient) (new pivot row)
Newz-row= current z-row(3) newpivot row
=(1, 2, 3, 0, 0, 0, 0) +
+3
_
0,
1
2
, 1,
1
4
, 0, 0, 5
_
=1,
1
2
, 0,
3
4
, 0, 0, 15
New s
2
-row = current s
2
-row(2) new pivot row
= (0, 2, 2, 0, 1, 0, 12)
2
_
0,
1
2
, 1,
1
4
, 0, 0, 5
_
=
= (0, 1, 0,
1
2
, 1, 0, 2)
New s
3
-row = current s
3
-row-(0) (new pivot row)
= current s
3
-row itself
= 0, 4, 0, 0, 0, 1, 16
Summarizing these results we get the newsimplex
tableau corresponding to the new basis (x
2
, s
2
, s
3
) as
follows. Note that this new basis corresponds to the
corner point E(0, 5) with value of OF as 15.
Basis

1
2
1
2
3
4
1
2
1
4
z
x
s
x
2
2
3
1
0
0
0
0
1
0
0
0
0
1
0
0
0
0
1
15
5
2
16
1
4 0
z x
1
x
2
s
1
s
2
s
3
solution
From the tableau, the solution is
x
2
= 5, s
2
= 2, s
3
= 16 (basic variables)
x
1
= 0, s
1
= 0 (non basic variables), value of OF is
15.
Thus the solution moved from corner point A to
corner point E in this one iteration. Optimal solution
is not reached since all elements of z-row are not
non negative. Since
1
2
is the most negative element
in the current z-row, the variable x
1
will enter the
basis. To determine the leaving variable again calcu-
late the ratios of RHS column with the elements of
the entering variable x
1
.
1
2
x
s
x
2
2
3
Basis Solution Ratio Entering
x
1
1
4
5
2
16
10
2
4
mini-
mum
Therefore, s
2
will leave the basis. The pivotal ele-
ment is one; so pivot row remain the same. The
chap-5a B.V.Ramana September 13, 2006 17:59
LINEAR PROGRAMMING 5.13
new simplex tableau corresponding to the new basis
(x
2
, x
1
, s
3
) is given below.
1
2
1
2
1
2
1
2
1
2
Basis Solution
z x
1
x
2
s
1
s
2
s
3
z
x
x
s
2
1
3
1
0
0
0
0
0
1
0
0
1
0
0
0
0
0
1
16
4
2
8

1
4
Here newz-row= current z-row
_

1
2
_
pivot row
=
_
1,
1
2
, 0,
3
4
, 0, 0, 15
_

1
2
__ _
0, 1, 0,
1
2
, 1, 0, 2
_
=
= 1, 0, 0,
1
2
,
1
2
, 0, 16
Here new x
2
-row = (current x
2
row)
1
2
(pivot row)
=
_
0,
1
2
, 0,
3
4
, 0, 0, 15
_

(1)
_
1
2
_
,
_
0, 1, 0,
1
2
, 1, 0, 2
_
=
=
_
1, 0, 0,
1
2
,
1
2
, 0, 16
_
Here new s
3
-row = current s
3
-row 4(pivot row)
= (0, 4, 0, 0, 0, 1, 16)
4
_
0, 1, 0,
1
2
, 1, 0, 2
_
=
= (0, 0, 0, 2, 4, 1, 8)
Since all the elements in the current z-roware non-
negative, the current solution is optimal. Read the
solution from the tableau as
x
2
= 4, x
1
= 2, s
3
= 8 (basic variables)
s
1
= 0, s
2
= 0 (non basic variables)
value of O. F is 16.
Note that this solution corresponds to the corner
point D(2, 4). In this second iteration solution moved
from E to D.
Simplex Method: Minimization:
Example 1: Minimize: z = x
1
+x
2
+x
3
subject
to x
1
x
4
2x
6
= 5, x
2
+2x
4
3x
5
+x
6
= 3,
x
3
+2x
4
5x
5
+6x
6
= 5.
Solution: Fortunately the problemcontains already
a starting basic feasible solution with x
1
, x
2
, x
3
as the
basic variables.
Basis Solution x
2
x
1
x
3
x
4
x
5
x
6
z
x
x
z
x
1
3
1
x
x
x
z
x
x
x
2
2
6
1
4
6
1
1
0
0
1
1
0
0
1
1
0
0
0
2
1
6
0
0
0
1
0
0
0
1
1
0
1
0
1
0
1
0
1
1
0
0
1
1
0
0
3
5
0
13
5
3
5
0
1
2
2
0
1
5
3
10
63
30
1
3
5
3
20
3
53
6
1
3
3
5
1
10
13
10
13
10
1
6
13
6
13
6
5
3
1
5
4
15
2
5
2
5
1
6
5
6
5
6
213
30
1
3
+

1 0 0
0
1
0
213
30
Optimal solution: x
1
=
213
30
, x
4
=
13
10
, x
6
=
2
5
O.F :
213
30
.
Unbounded solution:
Example 1: Solve LPP by simplex method.
Maximize: z = 2x
1
3x
2
+4x
3
+x
4
subject to
x
1
+5x
2
+9x
3
6x
4
2
3x
1
x
2
+x
3
+3x
4
10
2x
1
3x
2
+7x
3
8x
4
0
and x
1
, x
2
, x
3
, x
4
0.
Solution: Rewriting in the standard form
x
1
5x
2
9x
3
+6x
4
2
3x
1
x
2
+x
3
+3x
4
10
2x
1
+3x
2
7x
3
+8x
4
0
Introducting 3 slack variables x
5
, x
6
, x
7
we write the
LPP as
maximize: z = 2x
1
3x
2
+4x
3
+x
4
+0 x
5
+0
x
6
+0 x
7
subject to
x
1
5x
2
9x
3
+6x
4
+x
5
=2
3x
1
x
2
+x
3
+3x
4
+x
6
=10
2x
1
+3x
2
7x
3
+8x
4
+x
7
=0
chap-5a B.V.Ramana September 13, 2006 17:59
5.14 MATHEMATICAL METHODS
Objective equation is: z2x
1
+3x
2
4x
3
x
4
=0
The rst simplex tableau with the 3 basic variables
x
5
, x
6
, x
7
is given below:
Basis Solution x
2
x
1
x
3
x
4
x
5
x
6
x
7
z
x
x
x
5
6
7
2
1
3
2
3
5
1
3
4
9
1
7
1
6
3
8
0
1
0
0
0
0
1
0
0
0
0
1
0
2
10
0
Since 4 is most negative element in the z-row, the
associated variable x
3
will enter the basis. Out of the
three ratios
2
9
,
10
1
,
0
7
, the rst and third are ignored
(because the denominator is negative). So x
6
will
be outgoing variable. The pivotal element is 1. So
pivotal row remains same. The next simplex tableau
with x
5
, x
3
, x
7
is given below.
Basis Solution x
2
x
1
x
3
x
4
x
5
x
6
x
7
z
x
x
x
5
6
7
10
26
3
23
1
14
1
4
0
0
1
0
11
33
3
29
0
1
0
0
4
9
1
7
0
0
0
1
40
92
10
70
In the current z-row, x
2
has the most negative coef-
cient 1, so normally x
2
should enter the basis.
However, all the constraint coefcients under x
2
are
negative, meaning that x
2
can be increased inde-
nitely without violating any of the constraints. Thus
the problem has no bounded solution.
M-method:
Example 1: Solve the LPP by M-method
minimize z = 3x
1
+2.5x
2
subject to
2x
1
+4x
2
40
3x
1
+2x
2
50
x
1
, x
2
0
Solution: Introducing surplus variables x
3
, x
4
, the
greater than inequations are converted to equations.
Minimize z = 3x
1
+2.5x
2
+0 x
3
+0 x
4
subject to
2x
1
+4x
2
x
3
= 40
3x
1
+2x
2
x
4
= 50
x
1
, x
2
, x
3
, x
4
0
In order to have a starting solution, introduce two
articial variables R
1
and R
2
in the rst and second
equations. In the objective function the cost coef-
cients for these undesirable articial variables R
1
and
R
2
are taken as a very large penalty value M. Thus
the LPP takes the following form:
Minimize z = 3x
1
+2.5x
2
+0 x
3
+0 x
4
+
+ M R
1
+M R
2
(1)
subject to
2x
1
+4x
2
x
3
+R
1
= 40 (2)
3x
1
+2x
2
x
4
+R
2
= 50 (3)
and x
1
, x
2
, x
3
, x
4
, R
1
, R
2
0
The z-column is omitted in the tableau for conve-
nience because it does not change in all the iterations.
Solving (2) and (3) we get
R
1
= 40 2x
1
4x
2
+x
3
(4)
and R
2
= 50 3x
1
2x
2
+x
4
(5)
Substituting (4) and (5) in the objective function (1)
we get
z = 3x
1
+2.5x
2
+M(40 2x
1
4x
2
+x
3
)
+M(50 3x
1
2x
2
+x
4
)
or
z = (3 5M)x
1
+(2.5 6M)x
2
+M x
3
+
+M x
4
+90 M
which is independent of R
1
and R
2
. Thus the objec-
tion equation is
z (3 5M)x
1
(2.5 6M)x
2
Mx
3

Mx
4
= 90M
The simplex tableau with the starting basic solution
containing R
1
and R
2
as the basic variables in given
below:
Basis Solution x
1
x
2
x
3
x
4
R
1
R
2
z
R
R
1
2
3+5M
2
3
2.5 6M
4
2
M
1
0
M
0
1
0
1
0
0
0
1
90 M
40
50
In the z-row, the most positive coefcient is
2.5 +6M. So x
2
will be entering variable. Since
40
4
= 10,
50
2
= 25, the variable R
1
will leave the
chap-5a B.V.Ramana September 13, 2006 17:59
LINEAR PROGRAMMING 5.15
basis. So 4 is the pivotal element. New simplex
tableau is given below:
Basis Solution x
1
x
2
x
3
x
4
R
1
R
2
z
x
R
2
2
2
0
1
0
M
0
1
0
0
1
25+30M
40
50
3.5+4M
2
2.5 6M
4
2.5+2M
4
1
4
1
4
1
2
1
2
1
2

Since
35+4M
2
is the most positive element in the z-
row, the variable x
1
will enter the basis forcing R
2
out
since the minimum of the ratios
10
1
2
= 20,
30
2
= 15 is
15. So pivotal element is 2. The next simplex tableau
is shown below:
15
205
4
7
8
7
8
3
16
316 M
8
5
2
1
4
1
4
3
8
3
8
1
2
1
2
1
4
1
4
Basis Solution
x
1 x
2
x
3
x
4
R
1
R
2
M

z
x
x
2
1
0
0
1
0
1
0
Since all the elements in the z-row are non posi-
tive, the current solution is optimal given by x
1
= 15,
x
2
=
5
2
with value of objective function
205
4
(observe
that the articial variables R
1
, R
2
and surplus vari-
ables x
3
, x
4
are nonbasic variables assuming zero
values. Thus R
1
, R
2
have been forced out of the
basis).
Two-Phase Method
Example 1: Solve LPP by two-phase method
Maximize z = 2x
1
+3x
2
5x
3
subject to
x
1
+x
2
+x
3
= 7
2x
1
5x
2
+x
3
10
and x
1
, x
2
, x
3
0
Solution: Phase I: Introducing a surplus variable
x
4
and two articial variables R
1
and R
2
, the Phase
I of the LPP takes the following form:
Minimize r = R
1
+R
2
(1)
subject to
x
1
+x
2
+x
3
+R
1
= 7 (2)
2x
1
5x
2
+x
3
x
4
+R
2
= 10 (3)
and x
1
, x
2
, x
3
, x
4
, R
1
, R
2
0.
From(2), R
1
= 7 x
1
x
2
x
3
(4)
From(3), R
2
= 10 2x
1
+5x
2
x
3
+x
4
(5)
Substituting (4), (5) in (1) we get the objection func-
tion as
Minimize r = (7 x
1
x
2
x
3
) +(10 2x
1
+
5x
2
x
3
+x
4
)
or Minimize r = 3x
1
+4x
2
2x
3
+x
4
+17 or
r +3x
1
4x
2
+2x
3
x
4
= 17
The simplex tableau containing the basic solution
with R
1
, R
2
as the basic variables is given below.
Basis Solution x
2
x
1
x
3
R
1
R
2
x
4
r
R
R
1
2
3
1
2
4
1
5
2
1
1
0
1
0
0
0
1
1
0
1
17
7
10
The variable x
1
will enter the basis since 3 is most
positive coefcient in the r-row of this minimization
problem. The variable R
2
will leave the basis since
10
2
= 5 is less than
7
1
= 7. The pivotal element is 2.
Dividing the pivot rowby the pivot element 2, we get
the new pivot row as 1,
5
2
,
1
2
, 0,
1
2
,
1
2
, 5.
Here the new rth-row:
= (3 4 2 0 0 1 17) 3
_
1
5
2
1
2
0
1
2

1
2
5
_
=
_
0
7
2
1
2
0
3
2
1
2
2
_
Here the new R
1
-row:
= (1 1 1 1 0 0 7) 1
_
1
5
2
1
2
0
1
2

1
2
5
_
=
_
0
7
2
1
2
1
1
2
1
2
2
_
The new simplex table with R
1
and x
1
as the basic
variables is shown below:
Basis Solution x
2
x
1
x
3
R
1
R
2
x
4
r
R
R
1
2
0
0
1
0
1
0
2
7
5
7
2
1
2
3
2
1
2
7
2
1
2
1
2
1
2
1
2
1
2
1
2
5
2

Now x
2
with most positive coefcient
7
2
, will enter
chap-5a B.V.Ramana September 13, 2006 17:59
5.16 MATHEMATICAL METHODS
the basis pushing out R
1
with ratio
2
_
7
2
_
=
4
7
. (The
other ratio
5

5
2
is ignored since the denominator is
negative). The pivotal element is
7
2
. The pivot row is
_
0 1
1
7
2
7
1
7
1
7
4
7
_
Here new r-row:
=
_
0
7
2
1
2
0
3
2
1
2
2
_

7
2
_
0 1
1
7
2
7

1
7
1
7
4
7
_
= (0 0 0 1 1 0 0)
Here new x
1
-row:
=
_
1
5
2
1
2
0
1
2

1
2
5
_

5
2
_ _
0 1
1
7
2
7

1
7
1
7
4
7
_
=
_
1 0
6
7
5
7
1
7

1
7
45
7
_
The next simplex tableau of the second iteration with
x
1
and x
2
as the basic variables is given below.
Basis Solution x
2
x
1
x
3
R
1
R
2
x
4
r
x
x
2
1
0
0
1
0
1
0
0
1
0 0
1
7
2
7
1
7
1
7
4
7
6
7
5
7
1
7
1
7
45
7
1

The phase I is complete since r is minimized attain-


ing value 0, producing the basic feasible solution x
1
=
45
7
, x
2
=
4
7
. Note that both the articial variables R
1
and R
2
have been forced out of the (starting) basis.
Therefore the columns of R
1
and R
2
can altogether
be ignored in the future simplex tableau.
Phase II: Having deleted the articial variables
R
1
and R
2
and having obtained a basic feasible solu-
tion x
1
, x
2
we solve the original problem given by
maximization of z = 2x
1
+3x
2
5x
3
subject to
x
2
+
1
7
x
3
+
1
7
x
4
=
4
7
x
1
+
6
7
x
3

1
7
x
4
=
45
7
and x
1
, x
2
, x
3
, x
4
0
The tableau associated with this phase II is
Basis Solution x
2
x
1
x
3
x
4
z
x
x
2
1
3
1
0
2
0
1
0
1
7
1
7
90
7
12
7
102
7
4
7
6
7
1
7
45
7
5

+ =
Since x
2
with most negative element in the z-rowis
already in the basis, the current solution is optimal.
The basic feasible solution is x
1
=
45
7
, x
2
=
4
7
and
the maximum value of the objective function is
102
7
.
5.6 LINEAR PROGRAMMING PROBLEM
EXERCISE
Enumeration:
1. If a person requires 3000 calories and 100 gms
of protein per day nd the optimal product mix
of food items whose contents and costs are given
belowsuch that the total cost is minimum. Formu-
late this as an LPP. Enumerate all possible solu-
tions. Identify basic, feasible, nonfeasible, degen-
erate, non degenerate solutions and optimal solu-
tion.
Bread
x
1
Meat
x
2
Potatoes
x
3
Cabbage
x
4
Calories
Protein
Cost (Rs)
Milk
x
5
2500
80
3
3000
150
10
600
20
1
100
10
2
600
40
3
Ans: LPP: Minimize z = 3x
1
+10x
1
+x
3
+2x
4
+
3x
5
.
s.t. 2500x
1
+3000x
2
+600x
3
+100x
4
+
600x
5
= 3000
80x
1
+150x
2
+20x
3
+10x
4
+40x
5
= 100,
x
1
, x
2
, x
3
, x
4
, x
5
0; m= 2, n = 5,
5
c
2
= 10
basic solutions: F = Feasible, NF: non-feasible,
D: degenerate, ND: non degenerate
1. x
1
=
10
9
, x
2
=
2
27
, z =
110
7
, F, ND
2. x
1
= 0, x
3
= 5, z = 5, F, D
3. x
1
=
20
17
, x
4
=
10
17
, z =
80
77
, F, ND
4. x
1
=
15
13
, x
5
=
5
26
, z =
105
26
, F, ND, optimal,
5. x
2
= 0, x
3
= 5, z = 5, F, ND
6. x
2
=
4
3
, x
4
= 10, z =
20
3
, NF, ND
7. x
2
= 2, x
5
= 5, z = 5, NF, ND
8. x
3
= 5, x
4
= 0, z = 5, F, D
9. x
3
= 5, x
5
= 0, z = 5, F, D
10. x
4
= 10, x
5
= 30, z = 30, NF, ND
All the remaining non basic variables are
zero.
chap-5a B.V.Ramana September 13, 2006 17:59
LINEAR PROGRAMMING 5.17
2. Find all basic solutions for
x
1
+2x
2
+x
3
= 4, 2x
1
+x
2
+5x
3
= 5
Ans: (2, 1, 0) F, ND; (5, 0, 1), NF, ND;
_
0,
5
3
,
2
3
_
F, ND.
3. Find the optimal solution by enumeration
Max: z = 5x
1
+10x
2
+12x
3
s.t. x
1
, x
2
, x
3
0,
15x
1
+10x
2
+10x
3
200, 10x
1
+25x
2
+
20x
3
= 300
Ans: 1. (7.27, 9.1, 0, 0), z = 127 27
2. (5, 0, 12.5, 0), z = 175
3. (30, 0, 0, 250), NF
4. (0, 20, 40, 0), NF
5. (0, 12, 0, 80), z = 120
6. (0, 0, 15, 50), z = 180
(1) (2) (5) are F, ND:
(6) is optimal solution;
4. Find the optimal solution by enumeration
Max: z = 2x
1
+3x
2
s.t. 2x
1
+x
2
4, x
1
, x
2

0, x
1
+2x
2
5.
Ans: 1. (0, 0, 4, 5), z = 0, F, ND
2. (0, 4, 0, 3) NF
3. (0, 2.5, 1.5, 0), z = 7.5, F, ND
4. (2, 0, 0, 3), z = 4, F, ND
5. (5, 0, 6, 0), NF
6. (1, 2, 0, 0), z = 8, F, ND, optimal.
Simplex Method
Solve the following LPP by simplex method.
1. A rm can produce 5 different products using 3
different input quantities, as follows.
Input Technical coefcients Capacity
quantity 1 2 3 4 5
A 1 2 1 0 1 100
B 0 1 1 1 1 80
C 1 0 1 1 0 50
Prot 2 1 3 1 2
Maximize the prot
Ans: x
1
= 20, x
3
= 30, x
5
= 50, prot: Rs = 30
Hint: Max: z = 2x
1
+x
2
+3x
3
+x
4
+2x
5
s.t.
x
1
+2x
2
+x
3
+x
5
100; x
2
+x
3
+x
4
+x
5

80; x
1
+x
3
+x
4
50
2. Max: z = 2x
1
+x
2
s.t. x
1
, x
2
0; 3x
1
+5x
2

15; 6x
1
+2x
2
24.
Ans: x
1
=
15
4
, x
2
=
3
4
, z =
33
4
3. Max: z = 3x
1
+4x
2
+x
3
+7x
4
s.t. 8x
1
+3x
2
+
4x
3
+x
4
7,
2x
1
+6x
2
+x
3
+5x
4
3,
x
1
+4x
2
+5x
3
+2x
4
8
x
1
, x
2
, x
3
, x
4
0.
Ans: x
1
=
16
19
, x
4
=
5
19
, x
7
=
126
19
, z =
83
19
4. Minimize z = x
2
3x
2
+2x
5
s.t. x
1
+3x
2
x
3
+2x
5
= 7,
2x
2
+4x
3
+x
4
= 12,
4x
2
+3x
3
+8x
5
+x
6
= 10
Ans: x
2
= 4, x
3
= 5, x
6
= 11, z = 11
5. Max: z = 2x
1
+5x
2
+4x
3
s.t. x
1
+2x
2
+x
3
4; x
1
+2x
2
+2x
3
6
Ans: x
2
= 1, x
3
= 2, z = 13
6. Max: z = 5x
1
+4x
2
s.t., x
1
, x
2
0;
6x
1
+4x
2
24; x
1
+2x
2
6,
x
1
x
2
1; x
2
2
Ans: x
1
= 3, x
2
=
3
2
, z = 21
7. Max: z = x
1
+2x
2
+x
3
s.t. x
1
, x
2
, x
3
0,
2x
1
+x
2
x
3
2; 2x
1
+x
2
5x
3
6;
4x
1
+x
2
+x
3
6.
Ans: x
2
= 4, x
3
= 2, x
6
= 0, z = 10
(Note: Degenerate solution)
8. Max: z = x
1
+3x
2
2x
3
s.t.
3x
1
x
2
+2x
3
7, 2x
1
4x
2
12;
4x
1
3x
2
8x
3
10; x
1
, x
2
, x
3
0.
Ans: x
1
= 4, x
2
= 5, z = 11
9. Max. z = 6x
1
+9x
2
s.t. x
1
, x
2
0, 2x
1
+2x
2

24; x
1
+5x
2
44, 6x
1
+2x
2
60
Ans: x
1
= 4, x
2
= 8, x
6
= 20, z = 96
Multiple optima:
10. Minimize: z = x
1
x
2
s.t. x
1
, x
2
0
x
1
+x
2
2, x
1
x
2
1, x
2
1
Ans: x
1
=
3
2
, x
2
=
1
2
, z = 2
chap-5a B.V.Ramana September 13, 2006 17:59
5.18 MATHEMATICAL METHODS
Also another optimal solution is x
1
= 1, x
2
=
1, z = 2
11. Max: z = 6x
1
+4x
2
s.t. x
1
, x
2
0, x
1

4, 2x
2
12, 3x
1
+2x
2
18
Ans: x
1
= 4, x
2
= 3, z = 36
Another optimal solution: x
1
= 2, x
2
= 6, z =
36
Unbounded solution
12. Max: z = 4x
1
+x
2
+3x
3
+5x
4
s.t.
3x
1
2x
2
+4x
3
+x
4
10,
8x
1
3x
2
+3x
3
+2x
4
20,
4x
1
+6x
2
+5x
3
4x
4
20
Ans: Unbounded solution
Note: In the second simplex tableau, since x
2
has
most negative coefcient in z-row, normally x
2
should enter the basis. But all the entries in the
column under x
2
are negative or zero. So no vari-
able can leave the basis. Hence the solution is not
bounded
13. Min: z = 3x
1
2x
2
s.t. x
1
, x
2
0, x
1
x
2

1, 3x
1
2x
2
6.
Ans: Unbounded solution
Note: In the 3rd simplex tableau, x
3
having the
most positive value (12) in z-rowshould normally
enter the basis. But all the entries under x
3
are
negative. So OF can be decreased indenitely.
M-Method
14. Minimize: z = 4x
1
+2x
2
s.t. x
1
, x
2
0, 3x
1
+
x
2
27; x
1
x
2
21, x
1
+2x
2
30.
Ans: x
1
= 3, x
2
= 18, z = 48
15. Max: z = x
1
+2x
2
+3x
3
x
4
s.t.
x
1
+2x
2
+3x
3
= 15, 2x
1
+x
2
+5x
3
= 20,
x
1
+2x
2
+x
3
+x
4
= 10, x
1
, x
2
, x
3
, x
4
0
Ans: x
1
= x
2
= x
3
=
5
2
, x
4
= 0, z = 15
16. Min: z = 2x
1
+x
2
s.t. x
1
, x
2
0, 3x
1
+x
2
=
3, 4x
1
+3x
2
6, x
1
+2x
2
3
Ans: x
1
=
3
5
, x
2
=
6
5
, z =
12
5
17. Min: z = 3x
1
x
2
s.t. x
1
, x
2
0, 2x
1
+x
2

2; x
1
+3x
2
3; x
2
4
Ans: x
1
= 3, x
3
= 4, x
6
= 4, z = 9
18. Max: z = x
1
+5x
2
s.t. x
1
, x
2
0, 3x
1
+4x
2

6; x
1
+3x
2
2
Ans: x
2
=
3
2
, x
4
=
5
2
, z =
15
2
19. Min: z = 2x
1
+4x
2
+x
3
s.t. x
1
+2x
2
x
3

5; 2x
1
x
2
+2x
3
= 2; x
1
+2x
2
+2x
3
1
Ans: x
3
= 1, x
4
= 6, x
6
= 1, z = 1
Two-Phase Method:
20. Max: z = x
1
+5x
2
+3x
3
s.t. x
1
, x
2
, x
3
0, and
x
1
+2x
2
+x
3
= 3; 2x
1
x
2
= 4
Ans: (2, 0, 1), z = 5
21. Min: z = 4x
1
+x
2
s.t. x
1
, x
2
, x
3
, x
4
0, and
3x
1
+x
2
= 3; 4x
1
+3x
2
6, x
1
+2x
2
4.
Ans:
_
2
5
,
9
5
, 1, 0
_
, z =
17
5
22. Minimize z = 7.5x
1
3x
2
s.t. x
1
, x
2
, x
3

0, 3x
1
x
2
x
3
3; x
1
x
2
+x
3
2
Ans: x
1
=
5
4
, x
2
= 0, x
3
=
3
4
, z =
75
8
23. Minimize z = 3x
1
+2x
2
, s.t. x
1
, x
2
, 0, x
1
+
x
2
2; x
1
+3x
2
3, x
1
x
2
= 1
Ans: x
1
=
3
2
; x
2
=
1
2
, z =
11
2
24. Minimize: z = 5x
1
6x
2
7x
3
s.t. x
1
+5x
2

3x
3
15; 5x
1
6x
2
+10x
3
20; x
1
+x
2
+
x
3
= 5, x
1
, x
2
, x
3
0
Ans: x
2
=
15
4
; x
3
=
5
4
, x
5
= 30, z =
125
4
25. Max: z = 2x
1
+x
2
+x
3
s.t.
4x
1
+6x
2
+3x
3
8;
3x
1
6x
2
4x
3
1; 2x
1
+3x
2
5x
3
4;
x
1
, x
2
, x
3
0
Ans: x
1
=
9
7
; x
2
=
10
21
, z =
64
21
chap-5a B.V.Ramana September 13, 2006 17:59
LINEAR PROGRAMMING 5.19
5.7 THE TRANSPORTATION PROBLEM
The transportation problem is a special class of Lin-
ear programming problem. It is one of the earliest
and most useful application of linear programming
problem. It is credited to Hitchcock, Koopmans and
Kantorovich. The transportation model consists of
transporting (or shipping) a homogeneous product
fromm sources (or origins) to n destinations, with
the objective of minimizing the total cost of trans-
portation, while satisfying the supply and demand
limits.
Let a
i
denote the amount of supply at the ith
source, b
j
denote the demand at destination j; c
ij
denote the cost of transportation per unit from ith
source to jth destination; x
ij
the amount shipped
fromorigin i to destination j. Then the transportation
problem is to minimize the total cost of transporta-
tion
z =
m

i=1
n

j=1
c
ij
x
ij
(1)
subject to the constraints
source constraint :
n

j=1
x
ij
= a
i
,
a
i
> 0; i = 1, 2, . . . m (2)
Destination constraint:
m

i=1
x
ij
= b
j
, b
j
> 0; j = 1, 2, . . . , n (3)
and
x
ij
0 (4)
In the balanced transportation problem it is assumed
that the total quantity required at the destinations is
precisely the same as the amount available at the
origins i.e.
m

i=1
a
i
=
n

j=1
b
j
. (5)
(5) is the necessary and sufcient condition for the
existance of a feasible solution to (2) and (3).
Denoting the sources and destinations as nodes
and routes as arcs, the transportation problem can be
represented as a network shown below:
S
m
D
n
S
1
D
1
b
1
a
1
b
2
a
2
b
n
a
m
S
2
D
2
Demand Destinations Sources Supply
c x
11 11
;
c x
mn mn
;
The system of equations (1) to (4) is a linear pro-
gramming problemwith m+nequations in mnvari-
ables. The transportation problemalways has a nite
minimum feasible solution and an optimal solution
contains m+n 1 positive x
ij
s when there are m
origins and n destinations. It is degenerate if less than
m+n 1 of the x
ij
s are positive. No transportation
problem has ever been known to cycle.
Table for Transportation Problem
S
1
S
2
S
i
S
m
b
j
a
a
1
2
a
i
a
m
b
n
b
j
b
2
b
1
S = a
i
S b
j
c
1n
c
1j
c
12
c
11
c
mn
c
mj
c
m2
c
m1
x
1n
x
1j
x
12
x
11
x
mn
x
mj
x
m2
x
m1
K
K
a
i
D
n
D
j
D
2
D
1
K K
Destinations
S
o
u
r
c
e
s
Note: Zero values for nonbasic variables are not
lled while zero values for basic variables are shown
in the tableau.
Like the simplex method the transport algorithm
consists of determining the initial basic feasible solu-
tion, identifying the entering variable by the use of
optimality condition and nally locating the leaving
variable by the use of feasibility condition.
chap-5a B.V.Ramana September 13, 2006 17:59
5.20 MATHEMATICAL METHODS
Determination of Initial (starting) Basic Fea-
sible Solution
An initial basic feasible solution containing m+n
1 basic variables can be obtained by any one of the
following methods (a) the north west corner rule
(b) row minimum (c) column minimum (d) matrix
minimum (or least cost method) (e) Vogel approx-
imation method. In general, Vogels method gives
the best starting solution. Although computationally
north west corner rule is simple, the basic feasible
solution obtained by this method may be far from
optimal since the costs are completely ignored.
(a) North-west corner rule (due to Dantzig):
StepI: Allocate as much as possible to the north west
corner cell (1, 1). Thus let x
11
= min(a
1
, b
1
). If a
1

b
1
then x
11
= a
1
and all x
1j
= 0 for j = 2, 3, . . . n
i.e. except x
11
all other elements of the rst row are
zero. The rst row is satised so cross out the rst
row and move to x
21
of second row.
If a
1
b
1
then x
11
= b
1
and all x
i1
= 0 for i =
2, 3, . . . m i.e., except x
11
all other elements in the
rst column are zero. The rst column is satised
so cross out the rst column and move to x
12
of the
second column.
Note: If both a row and column are satised (i.e.,
say x
11
= a
1
= b
1
) simultaneously, then cross out
either row or column only but not both row and col-
umn.
Step II: Allocating as much as possible to the cell
(2, 1) or (1, 2) cross out the row or column and move
to (3, 1) or (1, 3).
Step III: If exactly one row or column is left
uncrossed out, stop. Otherwise go to step II wherein
move to lower row (below) if a row has just been
crossed out or move to right column if a column has
just been crossed out.
Note: Cells from crossed out row or column can
not be chosen for basis cells at a later step in the
determination of starting basic solution.
(b) Row-minimum
Identify the minimum cost element c
1k
in the rst
row. (Ties are broken arbitrarily). Allocate as much
as possible to cell (1, k). If a
1
b
k
then x
1k
= a
1
so
move to the second rowafter changing b
k
to b
k
a
1
.
Identify the minimum element in second row and
allocate as much as possible. Continue this process
until all rows are exhausted. If a
1
> b
k
then x
11
= b
k
,
change a
1
to a
1
b
k
, and identify the next smallest
(minimum) element in the rst rowallocate, continue
the process until the rst row is completely satised.
(c) Column-minimum
This is exact parallel to the above row-minimum
method except that minimum in the columns are
identied instead of rows.
(d) Matrix minimum (least-cost method):
Identify the least (minimum) element c
ij
in the
entire matrix. (Ties are broken arbitrarily). Allocate
as much as possible to the (i, j)th cell. If a
i
b
j
then x
ij
= a
i
, change b
j
to b
j
a
i
. If a
i
b
j
then
x
ij
= b
j
, change a
i
to a
i
b
j
. Identify the next least
element and allocate as much as possible. Continue
this process until all the elements in the matrix are
allocated (satised).
(e) Vogel approximation method
Step I. The rowpenalty for a rowis obtained by sub-
tracting the smallest cost element in that row from
the next smallest cost element in the same row. Cal-
culate the row penalties for each row and similarly
column penalties for each column.
Step II. Identify the row or column with the largest
penalty (ties are broken arbitrarily). In the selected
row or column, allocate as much as possible to the
cell withthe least unit cost. Cross out the satisedrow
or column. If a row and column are satised simul-
taneously, cross out either a row or column but not
both. Assign zero supply (or demand) to the remain-
ing row (or column). Any row or column with zero
supply or demand should not be used in computing
future penalties.
chap-5a B.V.Ramana September 13, 2006 17:59
LINEAR PROGRAMMING 5.21
Step III.
(a) A starting solution is obtained when exactly one
row or one column with zero supply or demand
remains uncrossed out. Stop.
(b) Determine the basic variables in an uncrossed
row (column) with positive (non zero) supply
(demand) by the least-cost method. Stop.
(c) Determine the zero basic variables in all the
uncrossed out rows and columns having zero sup-
ply and demand by the least cost method. Stop.
(d) Otherwise, go to step 1, recalculate the row and
column penalties and go to step II.
Note: Vogels method, which is a generalization of
the matrix minimum (least cost method) gives better
solution in most cases than all the other methods
listed above.
Method of Multipliers
The optimal solution to the transportation prob-
lem is obtained by iterative computions using the
method of multipliers (also known as UV-method
or stepping-stone method or MODI (modied distri-
bution) method). First of all, obtain a starting initial
basic feasible solution containing m+n 1 basic
variables (by any one of the above methods).
StepI: Introduce unknowns u
i
withrowi andv
j
with
column j such that for each current basic variable x
ij
in the tableau,
u
i
+v
j
= c
ij
is satised. This results in m+n equations in m+n
unknowns. Assume that u
1
= 1 (or u
1
= 0). (Instead
of u
1
, any other variable u
i
or v
j
can be chosen as
zero or one, resulting in the same optimal solution
but with different values in the tableau). Solving the
equations in u
i
, v
j
we get u
i
for i = 1 to m and v
j
for j = 1 to n.
Step II. For each non basic variable, compute
c
ij
= u
i
+v
j
c
ij
Step III: (a) If c
ij
0 for any i and j (i.e. for all
non basic variables), stop. The current tableau gives
the optimal solution with minimum cost.
(b) If c
ij
> 0, then solution is to be revised. The
entering variable is one which has most positive c
ij
(i.e., max c
ij
for all i and j).
(c) The leaving variable is determined by con-
structing a closed -loop which starts and ends at the
entering variable and consists of connected horizon-
tal and vertical lines (without any diagonals). Thus
each corner of the loop lies in the basic cell, except
the starting cell. The unknown is subtracted and
added alternatively at the successive corners so as
to adjust the supply and demand. From the cells in
which is subtracted, choose the maximum value
of such that x
ij
0. This feasibility condition
determines the leaving variable. Now go to step I.
Maximization A transportation problem in which
the objective is to maximize (the prot) can be trans-
fered to a minimization problem by subtracting all
the entries of the cost matrix from the largest entry
of the matrix.
Unbalanced problem in which the total supply is
not equal to the total demand can always be trans-
fered to a balanced transportation problem by aug-
menting it with a dummy source or dummy destina-
tion. A dummy destination is added when supply is
greater than the demand. The cost of transportation
from any source to this dummy destination is taken
as zero. Similarly when demand is greater than sup-
ply, a dummy source is added. The cost of shipping
from this dummy source to any destination is taken
as zero. Now the corresponding balanced problem is
solved by the method of multipliers.
Transhipment problem consists of transporting
from source to destination via (through) intermedi-
ate or transient nodes, known as transhipment nodes
which act as both sources and destination. The tran-
shipment node should be large enough to allow the
entire supply or demand to pass through it. Thus
the capacity of the transient node is the buffer
amount which equal the total supply or demand.
Thus the transhipment model consists of pure sup-
ply nodes which tranship the original supply, pure
demand nodes which receive the original demand,
and transhipment node which can receive original
supply plus the buffer or can tranship the original
demand plus the buffer. A given transhipment prob-
chap-5a B.V.Ramana September 13, 2006 17:59
5.22 MATHEMATICAL METHODS
lem can be transformed to a regular transportation
problem as follows:
I. Identify the pure supply nodes, pure demand
nodes and transhipment nodes fromthe given net-
work.
II. Denote the pure supply nodes and transhipment
nodes as the sources.
III. Denote the pure demand nodes and transhipment
nodes as the destinations.
IV. Note down the transportation costs c
ij
read from
the given network. If ith source is not connected to
jth destination, put c
ij
= M where M is a large
(penalty) value. Take c
ii
= 0 since it costs zero
for transporting from ith source to itself (ith des-
tination).
V. Identify supply at a pure supply node as the origi-
nal supply; demand at a pure demand node as the
original demand; supply at a transhipment node as
the sum of original supply and buffer and nally
demand at a transhipment node as the sum of the
original demand and buffer.
Nowthe above transformed regular transportation
problem can be solved by using the method of mul-
tipliers.
Degeneracy The solution of a transport problem
is said to be degenerate when the number of basic
variables in the solution is less than m+n 1. In
such cases, assign a small value to as many non-
basic variables as needed to augment to m+n 1
variables. The problem is solved in the usual way
treating the cells as basic cells. As soon as the
optimum solution is obtained, let 0.
5.8 Transportation Problem
WORKED OUT EXAMPLES
Starting Solution:
Example 1: Obtain a (non articial) starting basic
solution to the following transportation problem
using (a) North west corner rule (b) Row-minimum
(c) Column minimum(d) Least cost (Matrix minima)
(e) Vogels (approximation) method
S
S
S
1
2
3
0
2
1
4
3
2
2
4
0
D
1
D
2
D
3
7 6 6
8
5
6
Solution:
(a) NWCR:
S
S
S
1
2
3
0
2
1
4
3
2
2
4
0
D
1
D
2
D
3
7 6, 5 6
8, 1
5
6
7 1
5
6
Supply as much as possible to the north-west corner
cell (1, 1).
Cost: 7 0 +1 4 +5 3 +6 0 = 19
Note: This is a degenerate solution because it con-
tains only 4 basic variables (instead of 3 +3 1 = 5
basic variables).
(b) Row Minimum
S
S
S
1
2
3
0
2
1
4
3
2
2
4
0
D
1
D
2
D
3
7 6, 1 6, 5
8, 1
5
6
7 1
1
5
5 1
Allot as much as possible in the rst row to the cell
with least (minimum) cost i.e. (1, 1). The balance
allot to the next least cell in the rst row.
Cost: 7 0 +1 2 +5 3 +1 2 +5 0 =
19
Note: This is a non-degenerate solution (since it
contains 3 +3 1 = 5 basic variables).
chap-5a B.V.Ramana September 13, 2006 17:59
LINEAR PROGRAMMING 5.23
(c) Column Minimum
S
S
S
1
2
3
0
2
1
4
3
2
2
4
0
D
1
D
2
D
3
7 6, 1 6, 5
8, 1
5
6
7 1
6
5
Cost: 7 0 +1 2 +5 4 +6 2 = 34
This is a degenerate solution containing 4 basic
variables.
(d) Least cost method (matrix minima)
S
S
S
1
2
3
0
2
1
4
3
2
2
4
0
D
1
D
2
D
3
7 6, 1 6
8, 1
5
6
7 1
5
6
Allot as much possible to that cell which has least
cost in the entire matrix say (1, 1) (tie broken arbi-
trarily between (1, 1) and (3, 3).
Cost: 7 0 +1 4 +5 3 +6 0 = 19
This is a degenerate solution.
(e) Vogels (approximation) method
S
S
S
1
2
3
0
2
1
4
3
2
2
4
0
D
1
D
2
D
3
7 6 1
6
5
8 1
5
6
7
5
1
1 5
Row penalties
2
1
1
2
1
2
1
2
P
P
P
Column
penalties
1 1
1
1
2
2
4
1
Cost: 7 0 +2 1 +3 5 +2 1 +0 5 = 19
This is a non-degenerate solution.
Method of Multipliers
Example 1: Solve the following transportation
problem by UV-method obtaining the initial basic
solution by (a) Vogels method (b) NWCR (c) com-
pare the number of iterations in (a) and (b).
S
S
S
b
1
2
3
j
D
1
D
2
D
3
D
4
a
i
3
0
2
0
1
5
1
2
0
2
4
0
10 25 30 35 100
30
20
50
(a) Initial solution by Vogels method
P
PP
P
S
S
S
b
1
2
3
j
D
1
D
2
D
3
D
4
a
i
3
0
2
0
1
5
1
2
0
2
4
0
10 25
30 35
100
30
20
50
10
10
25 25 25
10 10
20 10
20 10
Row penalties
1
1
2
2
1
2
2
1
2
3
Column
penalties
2
2
3
1
1
1
1
2
2
2
2
1
Thus the initial basic feasible solution by Vogels
method is given by
S
S
S
1
2
3
D
1
D
2
D
3
D
4
3
0
2
0
1
5
1
2
0
2
4
0
10 10
25 25
20
where the basic variables are circled
Total cost: (20 0) +(10 2) +(10 0) +
(10 1) +(25 0) +(25 0) = 30
In the UV-method (method of multipliers) asso-
ciate the multipliers u
i
and v
j
with rowi and column
chap-5a B.V.Ramana September 13, 2006 17:59
5.24 MATHEMATICAL METHODS
j such that for each basic variable x
ij
we have
u
i
+v
j
= x
ij
Arbitrarily choosing u
1
= 1 we solve for the
remaining u
i
, v
j
s as follows:
Basic variable u, v equation solution
x
13
u
1
+v
3
= 0 v
3
= 1
x
14
u
1
+v
4
= 2 v
4
= 1
x
34
u
3
+v
4
= 0 u
3
= 1
x
32
u
3
+v
2
= 0 v
2
= 1
x
23
u
2
+v
3
= 1 u
2
= 2
x
21
u
2
+v
1
= 0 v
1
= 2
To summarize u
1
= 1, u
2
= 2, u
3
= 1
v
1
= 2, v
2
= 1, v
3
= 1, v
4
= 1
Now using u
i
and v
j
the non basic variables are
calculated as
x
ij
= u
i
+v
j
c
ij
Thus
Nonbasic Value
variable x
ij
u
i
+v
j
c
ij
x
11
u
1
+v
1
c
11
= 1 2 3 = 4
x
12
u
1
+v
2
c
12
= 1 +1 1 = 1
x
22
u
2
+v
2
c
22
= 2 +1 2 = 1
x
24
u
2
+v
4
c
24
= 2 +1 4 = 1
x
31
u
3
+v
1
c
13
= 1 2 2 = 5
x
33
u
3
+v
3
c
33
= 1 1 5 = 7
Non basic variables are placed in the south east
corner of each cell. Then the new table is
S
S
S
b
1
2
3
j
D
1
D
2
D
3
D
4
a
i
3
0
2
0
1
5
1
1
2
0
2
4
0
10 25 30 35
30
20
50
5
4
25 25
q + q
q
10
10
20
q
P
7
1
v
1
= 2 v
3
= 1 v
4
= 1 v
2
= 1
u
1
= 1
u
u
2
3
= 2
= 1
10
1
During computation, it is not necessary to write u,
v equations andsolve themexplicitly. Instead, choos-
ing u
1
= 1, compute v
3
, v
4
from the basic variables
x
13
, x
14
in the rst row. Now using v
4
, u
3
is obtained
from the basic variable x
34
. Similarly u
2
is obtained
using v
3
from the basic variable x
23
. Now using u
2
we get v
1
and nally using u
3
we get v
2
.
Incoming: Amongst the nonbasic variables, the
entering variable is the one with the most positive
value (in the south east corner of the cell). Thus x
21
will be the entering variable.
Outgoing: The leaving (basic) variable is deter-
mined by constructing a closed -loop which starts
and ends at the entering variable x
21
. In this modied
distribution all variables should be nonnegative and
supply and demand satised. Then
x
14
= 10 0
x
22
= 25 0
The maximum value of is 10 (which keeps both
x
14
, x
22
nonnegative i.e., x
14
= 0, x
22
= 15 > 0).
Thus the new table is
S
S
S
b
1
2
3
j
D
1
D
2
D
3
D
4
a
i
3
0
2
0
1
5
1
2
0
2
4
0
10 25 30 35
30
20
50
4
4
15 35
0
10 10
10 20
6
2
v
1
= 2 v
3
= 1 v
4
= 0 v
2
= 0
u
1
= 1
u
u
2
3
= 2
= 0
Since for all non basic variables x
11
, x
14
, x
22
, x
24
,
x
31
, x
33
the values (in the southeast) of u
i
+v
j
c
ij
are all negative, the current table is the optimal. The
optimal solution with least cost is (10 1) +(20
0) +(10 0) +(10 1) +(15 0) +(35 0) = 20.
(b) Initial solution by NWC rule: Suppressing
the working details we get the optimal solution in 3
iterations.
3
0
2
0
1
5
1
2
0
2
4
0
15 35
10
15
20
30
20
50
20
15
35
10 25 30 35
15 5
5
Associated cost
= (10 3) +(20 1) +(5 2) +(15 1)+
+ (15 5) +(35 0) = 150
chap-5a B.V.Ramana September 13, 2006 17:59
LINEAR PROGRAMMING 5.25
By UV method with u
1
= 1 we get
3
0
2
0
0 7
8
1
5
1
2
0
2
4
0
15 35
10
15 5
20
4
+

6 6
q
P
v
1
= 2 v
3
= 1 v
4
= 6 v
2
= 0
u
1
= 1
u
u
2
3
= 2
= 6
with = 5, x
22
, is outgoing and x
12
is the incom-
ing variable.
Then the new table is
3
0
2
0
6 1
8
1
5
1
2
0
2
4
0
10
35
10
20
5
20
2
+
6
0
q
P
2 5 0 0
1
4
0

with cost 120. Now choose = 10, x


13
will be
incoming and x
33
will be outgoing variable resulting
in the following table.
3
0
2
0
1
2
1
5
1
2
0
2
4
0
35
10
20
15
10 10
4 0
0
2 1 0 0
1
2
0

q
P
6
choose = 10. Then x
11
is outgoing and x
21
is
incoming with newfollowing table which is the opti-
mal solution since all x
ij
= u
i
+v
j
c
ij
0.
3
0
2
0
1 4
2
1
5
1
2
0
2
4
0
35
10
15
10 20
0
4
2 1 0 0
1
2
0
6
10
optimal cost is 20. Optimal solution is x
12
= 10,
x
13
= 20, x
21
= 10, x
23
= 10, x
32
= 15, x
34
= 35.
(c) The number of iterations is less when the
initial solution is obtained by Vogels method.
Unbalanced Transportation Problem
Example 1: Three electric power plants P
1
, P
2
, P
3
with capacities of 25, 40 and 30 kWh supply electric-
ity to three cites C
1
, C
2
, C
3
. The maximum demand
at the three cities are estimated at 30, 35 and 25 kWh.
The price per kWh at the three cites is given in the
following table
P
P
P
1
2
3
600
320
500
700
300
480
400
350
450
c
1
c
2
c
3
City
Plant
During the month of August, there is a 20%
increase in demand at each of the three cities, which
can be met by purchasing electricity from another
plant P
4
at a premium rate of Rs 1000/- per kWh.
plant 4 is not linked to city 3. Determine the most
economical plane for the distribution and purchase
of additional energy. Determine the cost of additional
power purchased by each of the three cities.
Solution:
P
P
P
1
2
3
P
4
600
320
500
1000
30+6
700
300
480
1000
35+7
400
350
450
M
25+5
25
40
70
13
148
c
1
c
2
c
3
City
Apply Vogels method to obtain the initial solution.
For all nonbasic variables u
i
+v
j
c
ij
0. The
present table is optimal. The optimal solution
is P
1
C
3
: 25, P
2
C
1
= 23, P
2
C
2
= 17, P
3
C
2
= 25,
P
3
C
3
= 5, P
4
C
1
= 13.
Total cost: Rs 36,710 + Rs 13000 = Rs 49710 only
city C
1
purchases an additional 13 kWh power from
plant P
4
at an additional cost of Rs 13,000/-.
chap-5a B.V.Ramana September 13, 2006 17:59
5.26 MATHEMATICAL METHODS
P
P
P
1
2
3
P
4
25
40
30
13
c
1
c
2
c
3
600
500
320
1000
700
480
300
1000
400
450
350
M
36 42 30
5 23
13
25
25
23 17
5
17
Row penalties
200
20
30
M
100
20
20
M
P
P
P P
Column
penalties
180
180
180
180
50
P
P
P
1
2
3
P
4
25
40
30
13
c
1
c
2
c
3
600
500
320
1000
700
480
300
1000
400
450
350
M
36 42 30
13
25
25
23 17
5
0
20 ve
151
80
u
1
= 0
u
u
u
2
3
4
= 129
= 51
= 551
v
1
= 449 v
2
= 429 v
3
= 399
271
Transhipment Problem
Example 1: The unit shipping costs through the
routes from nodes 1 and 2 to nodes 5 and 6 via nodes
3 and 4 are given in the following network. Solve the
transhipment model to nd how the shipments are
made from the sources to destinations.
1
2
3 5
4 6
100
200
1 6
5
8
4
3
2
3 1
150
150
Solution: The entire supply of 300 units is tran-
shipped from nodes P
1
and P
2
through T
3
and T
4
ultimately to destination nodes D
5
and D
6
. Here P
1
,
P
2
are pure supply nodes; T
3
, T
4
, D
5
are tranship-
ment nodes; D
6
is pure demand node. The tranship-
ment model gets converted to a regular transportation
problemwith 5 sources P
1
, P
2
, T
3
, T
4
, D
5
and 4 des-
tinations T
3
, T
4
, D
5
and D
6
. The buffer amount B =
total supply ( or demand) = 100 + 200 = (or 150 +
150) = 300 units. Ahigh penalty cost M is associated
with cell c
ij
when there is no route fromith origin to
the jth destination. Zero cost is associated with cells
(i, i) which do not transfer to itself. The initial solu-
tion is obtained by Vogels method. Taking u
1
=0 and
M = 99 and applying method of multipliers we have
P
P
T
T
D
1
2
3
4
5
99
1
4
4
1 99
2
99
1
2 3 99
99
92
1
92
0
90
6
8
90
0

3
2
5
92
1 99
196
99
195
0
150 150
q +q
0 0
q
300
300
200
100
100
200
300
300
300
u
1
= 0
u
u
u
u
2
3
4
5
= 2
= 91
= 0
= 97
T
3
T
4
D
5
D
6
v
1
= 1 v
2
= 0 v
3
= 97 v
4
= 98
P
300 300 450 150
q
value of objective function is Rs 2650. Not all c
ij

0. Note that c
43
= 92 is most positive so the variable
x
43
will enter into the basis. To determine the variable
leaving the basis, construct -loop from cells (4, 3)
to (2, 3) to (2, 2) to (4, 2). Choose = 0 to maintain
feasibility. Adjusting = 0, the leaving variable is
x
23
. Thus the new tableau is
p
p
t
t
d
1
2
3
4
5
99
93
4
4
1 99
94
99
91
2 3 99
92
99
92
1
0
0
2
6
8
2
0 3
2
5
1 99
104
99
103
0
150 150
+ q q
0
0
q
300
300
200
100
100
200
300
300
300
u
1
= 0
2
1
0
5
t
3
t
4
d
5
d
6
v
1
= 1 0 5 6
P
+ q q
q
300 300 150 150
value of objective function is Rs 2650. Not all
c
ij
0. Note that c
31
= 2 is most positive. So x
31
is the entering variable. To determine the leaving
chap-5a B.V.Ramana September 13, 2006 17:59
LINEAR PROGRAMMING 5.27
variable construct a -loop from cells (3, 1) to (2, 1)
to (2, 2), to (4, 2) to (4, 3) to (3, 3) to (3, 1). Choose
maximum value of = 200. Then x
21
will be the
leaving variable. Adjusting = 200, the new tableau
is given below.
p
p
t
t
d
1
2
3
4
5
99
91
4
2
1 99
92
99
91
2 3
2
99
92
99
92
1
0
0 6
8
2
0 3
4
5
1 99
104
99
105
0
150 150
100
200 100
200
200
100 100
200
300
300
300
0
0
1
2
7
t
3
t
4
d
5
d
6
1 2 7 8
300 300 450 150
Observe that all c
ij
0. Therefore the current
tableau is optimal. The basic feasible optimal
solution is x
11
=100, x
22
= 200, x
31
= 200, x
33
=
100, x
42
= 100, x
43
= 200, x
53
= 150, x
54
= 150.
The value of the objective function is
(100 1) +(200 2) +(200 0) +(100 6) +
(100 0) +(200 5) +(150 0) +(150 1) =
2250
Maximization:
Example 1: Solve the following transportation
problem to maximize the prot.
D
1
D
2
D
3
5
2
3
1
4
6
8
0
7
12
14
4
S
1
S
S
2
3
9 10 11
Solution: To transformthis problemto a minimiza-
tion, subtract all the cost entries in the matrix from
the largest cost entry 8. Then the relative loss matrix
is
P
3
6
5
7
4
2
0
8
1
12
14
4
3
2
1
9
2
10
2
11
1
Applying vogels method we get max penalty 3. So
allocate to (1, 3).
P
3
6
5
7
4
2
0
8
1
12
14
4
4
2
3
9
2
10
2
11
11 1
Allocate to (1, 1) since 4 is the largest penalty
P
3
6
5
7
4
2
0
8
1
12
14
4
4
2
3
9
2
10
2
11
11 1
Allocate to (3, 2) since 3 is the largest penalty.
P
3
6
5
7
4
2
0
8
1
1
14
4
2
3
9
8
10
2
11
11
1
1
3 3
6 6
5 5
7 7
4 4
2 2
0 0
8 8
1 1
14
8 6
11
1
4
8 6
4
1
Finally
11
The maximum prot (wrt the original cost
matrix) is (1 5) +(11 8) +(8 2) +(6 4) +
(4 6) = 157
chap-5a B.V.Ramana September 13, 2006 17:59
5.28 MATHEMATICAL METHODS
Degeneracy
Example 1: Solve the following TP using NWCR
0
2
1
4
3
2
2
4
0
8
5
6
S
1
S
S
2
3
7 6 6
D
1
D
2
D
3
Solution: By NWC rule
0
2
1
4
3
2
2
4
0
8
5
6
S
1
S
S
2
3
7 6 6
D
1
D
2
D
3
7 1
5
6
This is a degenerate solution since it contains only 4
basic variables (instead of 3 + 3 - 1 = 5 basic vari-
ables). To get rid of degeneracy, introduce any one
non-basic variable say in cell (3, 2) at level where
is a small quantity. Thus
0
2
3 3
3
1
4
3
2
2
4
0
S
1
S
S
2
3
D
1
D
2
D
3
7 1
5
e
6
1 3 1
0
u
1
= 1
0
1
since all c
ij
0 current solution is optimal. Now
letting 0 we get the solution as x
11
= 7, x
12
=
1, x
22
= 5, x
32
= 0, x
33
= 6 with OF = 19 +2 =
19 as 0
5.9 TRANSPORTATION PROBLEM
EXERCISE
1. Obtain the starting solution (and the correspond-
ing cost i.e. value of objective function: OF)
of the following transportation problems by (a)
North west corner rule (b) Rowminimum(c) Col-
umn minimum (d) Least cost method (e) Vogels
method.
1 5
0 2
3 3
2 1
4 4
1 6
6 8
2 0
5 7
7
12
11
12
14
4
10 9 10 10 10 11
(i) (ii)
10
4
13
14
3
20
15
9
7
12
5
7
12
1
5
7
9
8
0
19
(iii)
60 60 20 10
10
20
30
40
50
Ans. (i) (a) x
11
= 7, x
21
= 3, x
22
= 9, x
32
= 1, x
33
=
10, OF: 94
(i) (b) x
11
= 7, x
21
= 3, x
23
= 9, x
32
=
10, x
31
= 1, OF: 40
(i) (c) x
13
= 7, x
21
= 10, x
23
= 2, x
32
=
10, x
33
= 1, OF: 61
(i) (d) x
13
= 7, x
21
= 10, x
23
= 2, x
32
=
10, x
33
= 1, OF: 61
(i) (e) x
11
= 7, x
21
= 2, x
23
= 10, x
31
=
1, x
32
= 10, OF: 40
(ii) (a) x
11
= 9, x
12
= 3, x
22
= 7, x
23
=
7, x
33
= 4, OF: 104
(ii) (b) x
11
= 2, x
12
= 10, x
21
= 3, x
23
=
11, x
31
= 4, OF: 38
(ii) (c) x
12
= 10, x
13
= 2, x
21
= 9, x
23
=
5, x
33
= 4, OF: 72
(ii) (d) x
11
= 2, x
12
= 10, x
21
= 3, x
23
=
11, x
33
= 4, OF: 38
(ii) (e) x
11
= 2, x
12
= 10, x
21
= 3, x
23
=
11, x
33
= 4, OF: 38
(iii) (a) x
11
= 10, x
21
= 20, x
31
= 30, x
42
=
40, x
52
= 20, x
53
= 20, x
54
= 10, OF: 1290
(iii) (b) x
13
= 10, x
21
= 10, x
24
= 10, x
31
=
30, x
42
= 30, x
43
= 10, x
51
= 30, x
52
= 20, OF:
890
(iii) (c) x
13
= 10, x
21
= 20, x
31
= 10, x
33
=
10, x
34
= 10, x
42
= 40, x
51
= 50, OF: 860
chap-5a B.V.Ramana September 13, 2006 17:59
LINEAR PROGRAMMING 5.29
(iii) (d) x
12
= 10, x
22
= 20, x
31
= 10, x
32
=
20, x
42
= 10, x
43
= 20, x
44
= 10, x
51
= 50, OF:
960
(iii) (e) x
11
= 10, x
22
= 20, x
31
= 30, x
42
=
10, x
43
= 20, x
44
= 10, x
51
= 20, x
52
= 30, OF:
910
2. Solve the following TP by method of multipliers
method obtaining the starting solution by north
west corner rule.
(i) 3
0
2
3
0
0
2
5
4
0
4
1
1
1
0
3
1
1
6
0
30 60 50 40 20
40
70
60
30
Ans. x
21
= 40, x
21
= 10, x
22
= 20, x
24
= 40
x
31
= 20, x
33
= 40, x
43
= 10, x
45
= 20 OF:
70
(ii) 1
3
4
2
3
2
1 5
2 4
5 6
4 2
1 3
9 2
20 40 30 10 50 25
30
50
75
20
3 1 7 4 3 6
Ans. x
11
= 20, x
13
= 10, x
23
= 20, x
24
= 10, x
25
=
20
x
32
= 40, x
35
= 10, x
36
= 25, x
45
= 20 OF:
430
3. Solve the above problem 2 (ii) by obtaining the
initial solution by (a) row minimum (b) col-
umn minimum (c) matrix minimum (d) Vogels
method (e) compare the number of iterations
required in each of these methods including the
north west corner rule.
Ans. Optimal solution and value of OF is same as in 2
(ii) above for (a) (b) (c) (d). The number of itera-
tions required are 7 in NWCR, 1 in rowminimum
2 in column minimum, 1 in matrix minimum, 1
in Vogels method.
4. Solve the TP by UV-method obtaining initial
solution by
(a) NWC rule
Vogels method (c) compare the two meth-
ods.
10
12
0
0
7
14
20
9
16
11
20
18
5 15 15 10
15
25
5
Ans. (a) x
12
= 5, x
14
= 10, x
22
= 10, x
23
= 15, x
31
=
5 OF: 315
(b) same (c) Vogels method give solution
closer to optimal. Number of iterations required
in Vogel is one while 3 in NWC rule.
5. Solve the following TP by method of multipli-
ers by obtaining the starting solution by (a) NWC
rule (b) Least-cost method (c) Vogel approxima-
tion method. State the starting solution and the
corresponding value of OF.
10
12
4
2
7
14
20
9
16
11
20
18
5 15 15 15
15
25
10
S
1
S
S
2
3
D
1
D
2
D
3
D
4
Ans. Starting solution and associated OF (a) x
11
=
5, x
12
= 10, x
22
= 5, x
23
= 15,
x
24
= 5, x
34
= 10, OF: 520
(b) x
12
= 15, x
14
= 0, x
23
= 15, x
24
=
10, x
31
= 5, x
34
= 5, OF: 475
(c) x
12
= 15, x
14
= 0, x
23
= 15, x
24
=
10, x
31
= 5, x
34
= 5, OF: 475
(d) Solution by UV method
x
12
= 5, x
14
= 10, x
22
= 10, x
23
=
15, x
31
= 5, x
34
= 5, OF: 435
6. Solve the TP (use VAM)
21 25
17 14
16 13
18 23
6 10 12 15
11
13
19
32 18 27 41
D
1
D
2
D
3
D
4
S
1
S
S
2
3
chap-5a B.V.Ramana September 13, 2006 17:59
5.30 MATHEMATICAL METHODS
Ans. x
14
= 11, x
21
= 6, x
22
= 3, x
24
= 4, x
32
=
7, x
33
= 12 optimal minimum cost: Rs 796
Degeneracy:
7. Solve the following TP.
9
7
6
12
3
5
9 9
7 5
9 3
6 10
7 5
11 11
4 4 6 2 4 2
5
6
2
9
6 8 11 2 2 10
Ans. x
13
= 5, x
22
= 4, x
26
= 2, x
31
= 1, x
33
= 1,
x
41
= 3, x
44
= 2, x
45
= 4, x
13
=
minimum cost = 112 + 76 = 112 as 0
Hint: The starting solution obtained by Vogels
method is a degenerate since it contains only 8
basic variables (instead of 6 +4 1 = 9 basic
variables). Introduce any one of the non basic
variable at level where is small and let 0.
Maximization:
8. Solve the following TP to maximize the prot
15 42
80 26
51 33
42 81
23 31 16 30
23
44
33
90 66 40 60
Ans. x
12
= 23, x
21
= 6, x
22
= 8, x
24
= 30,
x
31
= 17, x
33
= 16, OF = 7005
Hint: Obtain the relative loss matrix by subtract-
ing all the entries of the cost matrix from the
largest entry 90.
Unbalanced TP:
9. Solve the following unbalanced TP.
11 7
21 10
20 8
16 12
30 25 35 40
50
40
70
8 18 12 9
D
1
D
2
D
3
D
4
S
1
S
S
2
3
Ans. x
13
= 25, x
14
= 25, x
23
= 10, x
25
= 30,
x
31
= 30, x
32
= 25, x
34
= 15
minimum cost = 1150
Hint: Total supply = 50 + 40 + 70 = 160
Total demand = 30 + 25 + 35 + 40 = 130
Introduce a dummy destination D
5
with
demand (requirement) of 30. Use VAM.
Note: x
25
= 30 means, 30 units are left
undespatched from S
2
. (Since it can not be send
to the dummy destination D
5
).
10. Solve the unbalanced TP;
5 7
6 6
1
4
75 20 50
10
80
15
3 5 2
D
1
D
2
D
3
S
1
S
S
2
3
Ans. x
12
= 10, x
21
= 20, x
22
= 10, x
23
= 50, x
31
=
15 minimum cost: 515
Hint: Introduce ctitious source S
4
with supply
of 40. Demand of 40 units is not met at destina-
tion 1.
Transhipment Problem
11. Solve the following transhipment problem
P
1
P
2
T
1
D
2
D
3
D
1
T
2
1000
1200
3
4
2
5
8
6
4
9
3
5
800
900
500
Ans. P
1
T
2
= 1000, P
2
T
1
= 1200, T
1
D
1
= 800
T
1
D
2
= 400, T
2
D
2
= 1000, D
2
D
3
= 500
minimum cost = (1000 4) +(1200
2) +(800 8) +(400 6) +(100 4) +
+(500 3) = 20, 700.
chap-5a B.V.Ramana September 13, 2006 17:59
LINEAR PROGRAMMING 5.31
Hint: The corresponding TP is given below:
3
2
0
M
M
M
4
5
7
M
0
M
M
M
8
0
M
M
M
M
6
5
4
0
M
M
M
M
9
3
P
P
T
T
D
D
1
2
1
2
1
2
T
1
T
2
D
1
D
2
D
3
1000
1200
B
B
B
B
B B 800+B 900+B 500
Here B = buffer = 1000 + 1200 = 2200 =
(800 + 900 + 500) and M = large penalty.
5.10 THE ASSIGNMENT PROBLEM
The assignment problem (or model) is a special case
of the transportation problemin which to each origin
there will correspond one and only one destination.
This can be described as a person-job assignment or
machine-task assignment model. Suppose there are
n persons who can perform any of the n different
jobs with varying degree of efciency measured in
terms of c
ij
representing the cost of assigning the i th
person to j th job (i = 1, 2, 3, . . . n, j = 1, 2, . . . n)
Then the objective of the assignment problem is to
minimize the total cost of performing all the njobs by
assigning the best person for the job on the one to
one basis of one person to one job. The assignment
problem can be solved as a regular transportation
problem in which the persons represent the sources,
the jobs represent the destinations, the supplyamount
at each source and demand amount at each destina-
tion being exactly equal to 1.
Let x
ij
=
_
1 if ith person assigned to j th job
0 if ith person not assigned to j th job
Since the ith person can be assigned to only one
job we have
n

i=1
x
ij
= 1 for i = 1, 2, . . . n.
Since the j th job can be assigned to only one
person we have
n

i=1
x
ij
= 1 for j = 1, 2, . . . , n. The
assignment problemconsists of determining the inte-
gers x
ij
(either 1 or 0) such that the total cost rep-
resented by the objective function
n

i=1
n

j=1
c
ij
x
ij
is
minimized. Thus the assignment problem is an inte-
ger linear programming problem. This combinato-
rial problem has n! number of possible assignments
which can be enumerated for small n. Even for n =
10, (n! = 3,628,800) the enumeration becomes very
time consuming and cumbersome. However the solu-
tion to the assignment problem is obtained by a sim-
ple method known as the Hungarian method or
Floods technique.
Hungarian Method or Floods Technique
1. Minimization case:
Step I. Determination of total-opportunity cost
(TOC) matrix:
(a) Subtracting the lowest entry of each column of the
given payoff (cost) matrix from all the entries of
that corresponding column results in the column-
opportunity cost matrix.
(b) Now subtracting the lowest entry of each row of
the column-opportunity matrix (obtained in step
(a) above) from all the entries of the correspond-
ing rowresults in the total opportunity cost (TOC)
matrix.
Step II. Check for optimal assignment: Let n be
the minimum number of horizontal and vertical lines
required which cover ALL the zeros in the current
TOC matrix. Let m be the order of the cost (TOC)
matrix.
(a) If n = m, an optimal assignment can be made.
Goto step V
(b) If n < m, revise the TOC matrix. Goto step III.
Step III. Revision of TOC matrix:
(a) Subtract the lowest entry (among the uncovered
cells) of the current TOC matrix from all the
uncovered cells.
(b) Add this lowest entry to only those cells at which
the covering lines of step II cross. This revises
TOC matrix.
chap-5a B.V.Ramana September 13, 2006 17:59
5.32 MATHEMATICAL METHODS
Step IV. Repeat steps II and III initial an optimal
assignment is reached:
Step V. Optimal assignment:
(a) Identify a rowor column (in the nal TOCmatrix)
having only one zero cell.
(b) Make assignment to this cell. Cross off both the
row and column in which this zero cell occurs.
(c) Repeat (a) for the remaining rows and columns
and make an assignment until a complete assign-
ment is achieved.
2. Maximization case:
The maximization problem can be converted to a
minimization problem by subtracting all the entries
of the original cost matrix from the largest entry (of
the original cost matrix). The transformed entries
give the relative costs.
3. Alternative Optima:
The presence of alternative optimal solutions is indi-
cated by the existance of a rowor column in the nal
TOC matrix with more than one zero cells.
4. Unbalanced Problem:
When the cost matrix is rectangular, a dummy row
or a dummy column added makes the cost matrix a
square matrix. All the costs c
ij
associated with this
dummy row (or column) are taken as zeros.
5. Problem with Restrictions:
When the assignment problem includes certain
restrictions such that a particular (specied) i th per-
son can not be assigned to a particular j th job then
the associated cost c
ij
is taken as a very big value M
(generally innity) so that it is prohibitively expen-
sive to make this undesirable assignment.
WORKED OUT EXAMPLES
Example 1: A national highway project consists
of 5 major jobs for which 5 contractors have submit-
ted tenders. The tender amounts (in lakhs of rupees)
quoted is given in the pay-off matrix below. If each
contractor is to be assigned one job, nd the assign-
ment which minimises the total cost of the project.
1
2
3
4
5
120
140
50
75
110
150
80
40
65
90
75
90
40
45
140
90
85
70
70
115
100
170
110
90
100
Cont-
ractor
Jobs
A B C D E
Solution: The column opportunity matrix is
obtained by subtracting the lowest entry in each col-
umn from all the entries in that column
1
2
3
4
5
70
90
0
25
60
110
40
0
25
50
35
50
0
5
100
20
15
0
0
45
10
80
20
0
10
A B C D E
Column-opportunity matrix
The total-opportunity-cost (TOC) matrix is now
obtained by subtracting the lowest entry from each
row from all the entries in that row.
1 60 100 25 10 0
2 75 25 35 0 65
3 0 0 0 0 25
4 25 25 5 0 0
5 50 40 90 35 0
A B C D E
Toc Matrix
Since the minimumnumber of vertical and horizontal
lines (n) needed to cover all the zeros is less than the
number of row m (or columns) i.e. n = 3 < m = 5,
the current TOCmatrix is to be revised by subtracting
the lowest entry among the uncovered cells from all
the uncovered cells and adding it to crossed cells
(where the vertical and horizontal lines intersect).
Thus the lowest entry 5 will be subtracted from all
chap-5a B.V.Ramana September 13, 2006 17:59
LINEAR PROGRAMMING 5.33
the uncovered cells and added at the crossed cells (3,
4), and (3, 5). Then the revised TOC is
1 55 95 20 10 0
2 70 20 30 0 65
3 0 0 0 5 25
4 20 20 0 0 0
5 45 35 85 35 0
A B C D E
Here the minimum number of lines covering all the
zeros is less than the number of rows i.e. n = 4 <
m = 5. Revise the matrix as above
1 35 75 0 10 0
2 50 0 10 0 65
3 0 0 0 25 45
4 20 20 0 20 20
5 25 15 65 35 0
A B C D E
Here n = 4 < m = 5
1 25 65 0 0 0
2 50 0 20 0 75
3 0 0 10 25 55
4 10 10 0 10 20
5 15 5 65 25 0
A B C D E
Optimal Toc matrix
n = 5 = m = 5
Ist optimal assignment: choose a row (or column)
containing only one zero. Choosing so the rst col-
umn, make an assignment. Thus
1 25 65 0 0 0
2 50 0 20 0 75
3 0 0 10 25 55
4 10 10 0 10 20
5 15 5 65 25 0
A B C D E

i.e. assign Ato 3. Cross off rst column and 3rd row.
Second and Third optimal assignment
1 25 65 0 0 0
2 50 0 20 0 75
3 0 0 10 25 55
4 10 10 0 10 20
5 15 5 65 25 0
A B C D E

Fromamong the remaining non-crossed out rows and


columns, choose a rowor column with only one zero.
Thus assign B to 2 and D to 1. Cross off the 2 row
and 2 column and 4th column and rst row.
4th and 5th optimal assignment:
1 25 65 0 0 0
2 50 0 20 0 75
3 0 0 10 25 55
4 10 10 0 10 20
5 15 5 65 25 0
A B C D E

i.e. assign C to 4 and E to 5


Thus the optimal assignment is
A3, B2, C4, D1, E5
with minimum cost = 50 + 80 + 45 + 90 + 100 = 365
Example 2: Maximization: The prot of assign-
ing a particular job to a specic machine is given in
the following matrix. Maximize the prot to accom-
plish all the jobs by assigning one machine to one
job. Check by enumeration.
1
2
3
380
210
260
610
380
210
330
415
300
A B C
Job
Machine
Solution: To convert into a minimization problem,
subtract all the entries of the matrix from the largest
entry 610. Then
1
2
3
230
400
350
0
230
400
280
195
310
A B C
chap-5a B.V.Ramana September 13, 2006 17:59
5.34 MATHEMATICAL METHODS
Job opportunity column matrix
1 0 0 85
2 170 230 0
3 120 400 115
A B C
TOC matrix
1 0 0 85
2 170 230 0
3 5 285 0
A B C
n = 2 < m = 3
Revised TOC matrix
1 0 0 90
2 165 225 0
3 0 280 0
A B C
n = 3 = m = 3
Optimal assignment: A3, B1, C2
1 0 0 90
2 165 225 0
3 0 280 0
A B C

Maximum prot: 260 + 610 + 415 = 1285


Check by enumeration:
A1, B2, C3 : 380 + 380 + 300 = 1060
A1, B3, C2 : 380 + 210 + 415 = 1005
A2, B3, C1 : 210 + 210 + 330 = 750
A2, B1, C3 : 210 + 610 + 300 = 1120
A3, B1, C2 : 260 + 610 + 415 = 1285 Optimal solu-
tion
A3, B2, C1 : 260 + 380 + 330 = 970
Example 3: Unbalanced problem: The amount of
time (in hours) to perform a job by different men
is given below. Solve the unbalanced problem by
assigning four jobs to three men subject to one job
to one man.
Men
M
M
M
1
2
3
J
1
7
5
8
J
2
5
6
7
J
3
8
7
9
J
4
4
4
8
Jobs
Solution: Add a ctitious (dummy) fourth man, to
convert the unbalanced to balanced assignment prob-
lem. The amount of time taken by the fourth man is
taken as zero for each job.
Men
M
M
M
1
2
3
M
4
J
1
7
5
8
0
J
2
5
6
7
0
J
3
8
7
9
0
J
4
4
4
8
0
Jobs
TOC matrix
3 1 4 0
1 2 3 0
1 0 2 1
0 0 0 0
J J J J
M
M
M
M
1 2 3 4
1
2
3
4
n m = 3 < = 4
Revised TOC Matrix
2 0 3 0
0 1 2 0
1 0 2 2
0 0 0 1
J J J J
M
M
M
M
1 2 3 4
1
2
3
4
= 4 = = 4 n m
Optimal assignment
2 0 3 0
0 1 2 0
1 0 2 2
0 0 0 1
M
1
J J J J
M
M
M
1 2 3 4
2
3
4

chap-5a B.V.Ramana September 13, 2006 17:59


LINEAR PROGRAMMING 5.35
M
1
J
4
, M
2
J
1
, M
3
J
2
, M
4
J
3
Thus the job J
3
is not assigned. The minimum
amount of time taken to accomplish all the three jobs
is 4 + 5 + 7 = 16 hours.
Persons
P
1
P
2
P
3
P
4
M
1
5
7
9
7
M
2
5
4
3
2
M
3

2
5
6
M
4
2
3

7
Machines
Example 4: Assignment with restrictions: The fol-
lowing matrix consists of cost (in thousands of
rupees) of assigning each of the four jobs to four
different persons. However the rst person can not
be assigned to machine 3 and third person can not be
assigned to machine 4.
(a) Find the optimal assignment to minimize the cost
(b) Suppose a 5th machine is available the assign-
ment costs to the four persons as 2, 1, 2, 8, respec-
tively. Find the optimal solution
(c) Is it economical to replace one of the existing
machines by the new (5th) machine
(d) In such case which machine is to be replaced
(unused). (A dash indicates that assignment is not
possible because of the restrictions imposed)
Solution: (a) Since rst person can not be assigned
to machine 3 a prohibitive (penalty) cost is imposed,
denoted by . Thus the cost matrix is
P
1
P
2
P
3
P
4
M
1
5
7
9
7
M
2
5
4
3
2
M
3

2
5
6
M
4
2
3

7
Row matrix
P
1
P
2
P
3
P
4
M
1
3
5
6
5
M
2
3
2
0
0
M
3

0
2
4
M
4
0
1

5
TOC Matrix Revised TOC
0
2
3
2
0
1
2
1

0
2
4

0
2
4
3
2
0
0
4
2
0
0
0
1
5

0
0
4

n = 3 < m = 4 n = 3 < m = 4
Optimal assignment
0 5 0
1 3 0 0
1 0 1
0 0 3 3
M
P
1

1 2 3 4
2
3
4
M M M
P
P
P

P
1
to M
4
, P
2
to M
3
, P
3
to M
2
, P
4
to M
1
minimal cost: 2 + 2 + 3 + 7 = 14
(b) Introducing the 5th machine, the new cost matrix
is
P
P
1
5
P
P
P
2
3
4
M
1
3
7
9
7
0
M
2
5
4
3
2
0
M
3

2
5
6
0
M
4
2
3
7
0

M
5
2
1
2
8
0
Subtracting the row minimums from each row, we
get

3 3 0 0
6 3 1 2 0
7 1 3 0
5 0 4 5 6
0 0 0 0 0

L
N
M
M
M
M
M
M
O
Q
P
P
P
P
P
P
3
5
6
4
0
4
3
1
0
1

0
2
3
0
0
1
4
0

0
0
0
5
1
M
1
P
1
P
P
P
P
2
3
4
5
M
2
M
3
M
4
M
5
n m = 4 < = 5 n m = 5 = 5
optimal matrix
Optimal assignment:
P
1
to M
4
; P
2
M
3
, P
3
M
5
, P
4
M
2
; P
5

M
1
Minimum cost: 2 + 2 + 2 + 2 = 8
(c) With the introduction of 5th machine the cost
has come down from 14 to 8. So it is economical to
introduce 5th machine.
chap-5a B.V.Ramana September 13, 2006 17:59
5.36 MATHEMATICAL METHODS
(d) Since the dummy person P
5
is assigned to
machine 1, it means that M
1
is not used and therefore
can be replaced (dispensed with).
5.11 ASSIGNMENT PROBLEM
EXERCISE
1. Solve the following assignment Problemfor min-
imum total cost. Check by enumeration.
1
2
3
A
380
210
260
B
610
380
210
C
330
415
300
Job
Machine
Ans. A
2
, B
3
, C
1
: minimal cost: 750
A
1
B
2
C
3
(1060), A
1
B
3
C
2
(1005), A
2
B
1
C
3
(1120),
A
2
B
3
C
1
(750), A
3
B
1
C
2
(1285), A
3
B
2
C
1
(970).
2. Find the optimal solution with minimum cost in
the following assignment problem
O
O
O
1
2
3
20
10
14
27
18
16
30
16
12
Origins D
1
D
2
D
3
Destinations
Ans. O
1
to D
2
, O
2
to D
1
; O
3
to D
3
Total minimum cost: 49
3. Find the optimal assignment which maximizes
the total cost in the above problem 2.
Ans. O
3
to D
1
, O
2
to D
2
, O
1
to D
3
, maximumcost: 62
Solve the following assignment problem for
minimum cost
4.
P
P
P
1
2
3
15
9
10
10
15
12
9
10
8
Person J
1
J
2
J
3
Jobs
Ans: P
1
to J
2
, P
2
to J
1
, P
3
to J
3
, minimum cost: 27
5.
P
P
P
1
2
3
P
4
1
9
4
8
4
7
5
7
6
10
11
8
3
9
7
5
Persons J
1
J
2
J
3
J
4
Jobs
Ans. P
1
to J
1
, P
2
to J
3
; P
3
to J
2
, P
4
to J
4
minimum
cost: 21
6. Assign the ve jobs to the ve machines so as to
maximize the total return if the following matrix
shows the return in (thousands of) 1 rupees for
assigning the ith machine (i = 1, 2, 3, 4, 5) to the
jth job (j = 1, 2, 3, 4, 5).
Machine 1 2 3 4 5
1 5 11 10 12 4
2 2 4 6 3 5
3 3 12 5 14 6
4 6 14 4 11 7
5 7 9 8 12 5
Job
Ans. M
1
J
3
; M
2
J
5
; M
3
J
4
; M
4
J
2
; M
5
J
1
, Maximum
cost: 50
7. Four men can perform any of the four tasks with
different efciency measured in terms of time
required to complete each task which is given in
the following table. Assign one task to one man
so as to minimize the total time spent on accom-
plishing the four tasks.
Men
Task
18 26 17 11
13 28 14 26
38 19 18 15
19 26 24 10
M
T
T
T
T
1 2 3 4
1
2
3
4
M M M
Ans. M
1
to T
2
; M
2
to T
3
; M
3
to T
1
; M
4
to T
4
, minimum
total time = 59
chap-5a B.V.Ramana September 13, 2006 17:59
LINEAR PROGRAMMING 5.37
8. Multiple Optima: Solve the following assignment
problem for minimum cost.
20 13 7 5
25 18 13 10
31 23 18 15
45 40 23 21
M
M
M
M
1
2
3
4
Jobs J J J J
1 2 3 4
Men
Ans. Four optimal assignments all with the same min-
imum total cost: 76.
(i) M
1
J
1
; M
2
J
4
; M
3
J
2
; M
4
J
3
.
(ii) M
1
J
2
; M
2
J
1
; M
3
J
4
; M
4
J
3
.
(iii) M
1
J
1
; M
2
J
2
; M
3
J
4
; M
4
J
3
.
(iv) M
1
J
4
; M
2
J
1
; M
3
J
2
; M
4
J
3
.
9. Unbalanced assignment problem Three work
centers are requiredtomanufacture, assemble and
to package a product. The handling cost at each
of the four locations in the factory are given in
the following matrix. Determine the location of
work centres that minimizes total handling cost.
Job
Manufacturing 18 15 16 13
Assembly 16 11 15
Packaging 9 10 12 8
L L L L
M
M
M
1 2 3 4
1
2
3
Locations
Ans: M
1
to L
4
; M
2
to L
2
; M
3
to L
1
. Location 3 is kept
idle (assigned to dummy job and no job is done).
10.
A B C D E
M
M
M
M
1
2
3
4
62 78 50 101 82
71 84 61 73 59
87 92 111 71 81
48 64 87 77 80
Jobs
M
e
n
Maximize the prot. Which job should be
declined?
Ans. M
1
D; M
2
B; M
3
C; M
4
E, M
5
A. Since (dummy
man) M
5
is assigned to job A, the job A should
be declined. Maximum prot: 101 + 84 + 111 +
80 = 376
11. With Restrictions Suppose ve men are to be
assigned to ve jobs with assignment costs given
in the following matrix. Find the optimal assign-
ment schedule subject to the restriction that rst
person M
1
can not be assigned job 3 and third
person M
3
can not be assigned to job 4.
Jobs
Men
5 5 2 6
7 4 2 3 4
9 3 5 3
7 2 6 7 2
6 5 7 9 1
J J J J J
M
M
M
M
M
1 2 3 4 5
1
2
3
4
5
Ans. Minimal cost is 15 with three alternative optimum
solutions.
(i) M
1
J
4
, M
2
J
3
, M
3
J
2
, M
4
J
1
, M
5
J
5
(ii) M
1
J
4
, M
2
J
3
, M
3
J
5
, M
4
J
2
, M
5
J
1
(iii) M
1
J
4
, M
2
J
3
, M
3
J
2
, M
4
J
5
, M
5
J
1
12. Determine optimal location of three machines at
four different locations in a shop oor given the
cost estimate per unit of time of material han-
dling is given in the following matrix. Note that
machine 2 can not be placed in location 2.
Machines 2 3 4
M1 12 9 12 9
M2 15 - 0 20
M3 4 8 115 6
L1 L L L
Locations
Ans. M
1
to L
2
or L
4
; M
2
to L
3
, M
3
to L
1
Dummy (ctitious) machine M4 is assigned
to location L
4
or L
2
.
Minimum cost: 9 + 13 + 4 = 26