INTEGER PROGRAMMING

ntroduotion

An integer programming model is one where one or more of the decision variables has to take on
an integer value the final solution. Solving an integer programming problem is much more
difficult than solving an LP problem. Even the fastest computers can take an excessively long
time to solve big integer programming problems. If requiring integer values is the only way in
which a problem deviates from a linear programming formulation, then it is an integer
programming (|P) problem. (1The more complete name is integer linear programming, but the
adjective linear nomally is dropped except when this problem is contrasted with the more
esoteric integer nonlinear programming problem. So, the mathematical model for integer
programming is the linear programming model with the one additional restriction that the
variables must have integer values.
In linear programming, each of the decision variable as well as slack and lor surplus variable is
allowed to take any fractional value. However, there are certain practical problems in which the
fractional value of the decision variables has no significance. For example, it does not make
sense saying 1.5 men working on a project or 1.6 machines in a workshop. The integer solution
to a problem can, however, be obtained by rounding off the optimum value of the variables to the
nearest integer value.

This approach can be easy in terms of economy ofeffort, time and cost that might be required to
derive an integer solution but this solution may not satisfy all the given constraints. Secondly, the
value of the objectivefunction so obtained may not be optimal value. All such difficulties can be
avoided if the given problem, where an integer solution is required, is solved by integer
programming techniques.
Why used?

+ The inmplicit assumption was that solutions could be fractional or real numbers (i.e.non
integer) in the linear programming models. However, non-integer solutions are aways
practical.
+ When only integer solutions are practical or logical, non-integer solution values can be
rounded off to the nearestfeasible integer values.
+ For example, if the case is nail, it will cause little concern considering 8000.4 nails as
8000 nails. Round off makes the case cost only afew cents apiece.
+ For example. if the case is a production of jet aircraft, round off could affect profit or
cost by million dollars considering 7.4 jet airliners as 7 or &.

Integer Programming Models

There are three types of models;
cOMPILED BY TSEGAY BERHE MSc IN PRODUCTION ENGINEERING AND MANAGEMENT
 V Pure Integer Model; All decision variables required to have integer solution values.
0-1 Integer Model: All decision variables required to have integer values of zeroor one.
Mixed Integer Model: Some of the decision variables (but not all) required to haw
integer values.

There are certain decision problems where decision variables make sense only if they have
integer values in the solution. Capital budgeting, construction scheduling. plant location and
size, routing and shipping schedule, batch size, capacity expansion, fxed charge, etc., are few
problems which demonstrate the areas ofapplication of integer programming.
Example 1:A Total (Pure) Integer Model
Machine shop obtaining new presses and lathes. The Marginal profitability of each press
machine is S100/day and each of lathe machine is S150lday. The Resource constraints is
S40,000 and 200 sq. fi. floor space. The Machine purchase prices and space requiements are
given in the table below:
Required Floor Space (sq.
Machine.) Purchase Price
Press 15 8000
Lathe 30 4000

Integer Programming Model:

Maximize Z= $100X+ S150X;

subject to:
8,000X,+ 4,000X, < S40,000

15X,+ 30X,s 200f

X, X2>0 and integer

X= number of presses
X= number of lathes

Example 2: A Total Integer Model

Integer Programming Model:
Maximize Z = 6X}+ 5X, + 2X;

cOMPILED BY TSEGAY BERHE MSCIN PRODUCTION ENGINEERING AND MANAGEMENT 2

 subject to:

10X;+ 4X+ 2 X;s600

2X,+ 5 X,+ 2 X; < 800

X, X, X, >0 and integer

X= 22, X,= 0, X, = 378

Z-888

Example 3: A0- 1 Integer Model

Recreation facilities selection to maximize daily usage by residents.

Resource constraints: SI20,000 budget; 12 acres of land.

Selection constraint: either swimming pool or tennis center (not both).
Data:
Expected
Recreation Usage Cost Land
Facility (people/day) (S) Requirement
(acres)
Swimming
pool 300 35000 4
Tennis Center 9 10000 2
Athletic field 400 25000 7
Gymnasium 150 90000 3

Integer Programming Model:

Maximize Z= 300X+ 90X;+ 400X,+ 150x4
subject to:
$35,000X+ 10,000X,+ 25,000X;+ 90,000X,S S120,000

4X,+ 2X;t 7Xs+ 3XsI2 acres

X+ X,s lfacility
X, X, Xs, X = 0 or 1

cOMPILED BY TSEGAY BERHE MSC IN PRODucTION ENGINEERING AND MANAGEMENT

 X= construction of aswimming pool
X= construction ofa tennis center

X= construction of anathletic field

X= construction of agymnasium

Example 4:

Integer Programning Model:

MaximizeZ = 6X,+ 5X, + 2X;
subject to:
10X;+ 4X+ 2 X,<600
2X;+ 5 X+ 2X,< 800
XI, X2, X3 >0 OR Il

X,= 1, X,= 1, X, = 1, Z=13

Example 5: A Mixed Integer Model:

S250,000 available for investments providing greatest return after one year.
Data:

Condominium cost S50,000/unit, $9,000 profit if sold one year.

Land cost SI2,000/ acre, S1,500 profit if sold after oneafter
year.
4 Municipal bond cost
$8,000/bond, S1,000 profit if sold after one year.
Only 4 condominiums, 15 acres of land, and 20
municipal bonds available.
Integer Programming Model:
Maximize Z = $9,000x1+ 1,500x2 + 1,000x3
subject to:

50,000X, + 12,000X, + 8,000X, < S250,000

cOMPILED BY TSEGAY BERHE MSC IN PRODUCTION

ENGINEERING AD MANAGEMENT
 X,s4 condominiums

X,<15 acres
Xs20 bonds

X; >0

X, X, >0 and integer

Xj= condominiums purchased, X, = acres of land purchased, X= bonds purchased

Example 6:

Integer Programming Model:

Maximize Z = 6X)+ 5X, + 2X,

subject to:

10X,+ 4X,t 2 X, S600

2X,+ 5X;+ 2X; S800
X, X, X; 20
X,= 22.2, X,= 0, X, = 377.8

Z=888.9

METHODS FOR SOLVING ILP PROBLEMS

a. Branch-and-Bound Method (Developed By: A.H. Land and A. G. Doing)

b. Cutting-Plane Method (developed by: Ralph E. Gomory)
a. BRANCH-AND-BOUND METHOD

Creates and solves a sequence ofsub-problems to the original problem that are increasingly
more restrictive until an optimal solution is found.

Branching: Selection ofan integer value of a decision variable to examine for a possible
integer solution to aproblem. "If the solution to the linear programming problem contains
non-integer values for some or all decision variables, then the solution space is reduced by
introducing constraints with respect to any one ofthose decision variables. Ifthe value ofthe
cOMPILED BY TSEGAYBERHE MSC IN PRODUCTION ENGINEERING AND MANAGEMENT
 decision variable "X1" is 2.5, then hso more problems will be created by using each of the
following constraints. XI<2 andXI 2 3.

Bound: An upper or lower limit on the value of the objectivefunction at a given stage of th
analysis of an integer programming problem.
Lower bound: The lower bound at a node is the value of the objective finction
corresponding to the truncated alues (integer parts) of the decision variables of the problem
in that node.

Upper bound: The upper bound at a node is the value of the objective function
corresponding to the linear programming solution in that node.

Fathomed subproblem /node: Aproblem is said to be fathomed if any one of the following
three conditions is true:

1. The values of the decision variables of the problem are integer.

2. The upper bound of the problem which has non-integer values for its decision variables
is not greater than the current best lower bound.
3. The problem has infeasible solution. This means that further branching from this type of
fathomed nodes is not necessary.

Current best lower bound: This is the best lower bound (highest in the case ofmaximization
problem and lowest in the case of minimization problem) among the lower bounds of all the
fathomed nodes. Initially, it is assumed as infinityfor the root node.
Branch &Bound Algorithm Applied to Maximization Problem:
Step 1. Solve the given linear programming problem graphically or using iterative
method. Set, the current best lower bound ZB as o.
Step 2. Check, Whether the problem has integer solution. Ifyes, print the current solution
as the optimal solution and stop: Otherwise go to Step-3.
Step 3. ldentify the variable Xk which has the maximum fractional part as the branching
variable. (In case of tie, select the variable which has the highest objective function
coefficient.)
Step 4. Create two more problems by including each of the following constraints to the
current problem and solve them.
a) XsInteger part ofX
b) Xi >Next Integer of X
Step 5. If any one of the new sub-problems has infeasible solution orfuly integer values
for the decision variables, the corresponding node is fathomed. Ifa new node has
integer values for the decision variables, update the current best lower bound as the
lower bound of that node if its lower bound is greater than the previous current best
lower bound.

cOMPILED BY TSEGAY BERHE MSC IN PRODUCTION ENGINEERING AND MANAGEMENT 6

 Step 6. Are all terminal nodes fathomed? If answer is yes, go to step-7; otherwise,
identify the node with the highest lower bound and go to step-3.
Step 7. Select the solution of the problem with respect to the fathomed node whose lowe.
bound is equal to the current best lower bound as the optimal solution.
Example
Max. Z 10X, + 20X,
Subject to:
6X, + 8X,<48
X, + 3X, sI2
XI, N, 2 0 and integers
Example 1
Z(A)=10*0+20"0=0
6XI + 8N2 s 48 ZB)=10*8+20"0=$0

6
Z(C)=10*(24/5(+20*(12/5)-96
Z(D)=10"0+-20*4=80

2
C24/5,12/5)
1 XI + 3X? s 12

2 4 6 8 10 12 14

Zy= Upper bound = Z (Optimum) ofLP Problem.

Z Lower bound w. r. t. the truncated values of the decision variables
Z=Current Best Lower Bound
X_24/5, X;=12/5,Z=96

Zy=96
Z;=80

In Problem (P1),. X, has the highest fractional part 24/5. Hence; "X," is selected for further
branching.

cOMPILED BY TSEGAY BERHE MSC IN PRODUCTION ENGINEERING AND MANAGEMENT

cOMPILED BY TSEGAY BERHE MSC IN PRODUCTION ENGINEERING AND MANAGEMENT

Max. Z = 10X, + 20X,

Subject to:
6x, +8X,s48
X + 3X, s12
X, X; 2 0 and integers
X25 Xs4
P, P
solution Max. Z= 10X, + 20X, solution Max Z = 10X, + 20X,
X-5 Subject to: X4 Subject to:
X-94 6X, + 8X,s48 X=83 6x, +SX, s48
Zu-95 X +3X,s12 Zu-93.33 X+ 3X, s12
ZË=90 Z_=80
X, X,0 and integers X, X, 20 and integers

The Problem P2 has the highest lower bound (Z) of 90 among the unfathomed ternminal nodes.
So, the further branching is done from this node.
 Max Z=10X;+ 20X,
by tsegay b |Subject to:
6X, +8X,S48
X+ 3X, <12
XX,20 and integers

X25 Xs4
P; P
solution Max Z=10X, +20X, solution Max Z=10X, +20X,
|Subject to: Subject to:
K-94 K-83 6X, +SX, s48
Z-95 K +3X,s12 Z-9333 K+3X;s 12
Z-80
K, X,20 nd integers XX20and integers
X23
P P
Solution Max. Z=10X, +20X, Solution Max Z=10X, +20X,
Infeasible sohution Subject to: X=163 |Subject to:
6X, +8X, s 48 6X, +8X,S48
X+3X,s 12 Zu-9333
|X25 Z=90
XI23 XI2
XX,>0 and integers K, X,20 and integers

COMPILED BY TSEGAY BERHE MSC IN PRODUCTION ENGINEERING AND MANAGEMENT

 Max. Z=10X, +20X,
by tsegay b Subject to:
|6X, +8X, s48
X, +3X,s 12
X,X,20 and integers

X25 Xs4
P
solution Max Z=10X, +20X, solution Max. Z=1OX, 20X,
Subject to:
X-94 6X, +8X,s 48
Subject to:
Zu-95
Ky=83 6X, +8X,s 48
X+3X,S 12 Zu-93 33
Z=90 K, +3X; s12
Zz=80 Xs4
XX,20 and ztegers
X.X,20 and integers
XIs2
P P
Solution Max. Z= 10X,+20X;
Infeasible soltion
Solution Max. Z=10X, +20X,
|Subject to: X=163
fathomed Subject to:
6x, +SX,s48 X;=2 |6x, +8X,s48
X+3X;s 12 Zu-93 33 K, +3X;s 12
X25 Zz-90
Xl23
X,X,20 and integers
XX,>0 and integers
X26
P
Solution P
Max Z= 10X,+20X;
X=6
Solution Max. Z= 10X, +20X,
|Subject to:
X,=3/6 6X, +8X,s 48 Subject tor
Zu-90 X=2 |6x, +8X, s48
,+3X;s 12 Zu-90
Z=80 X +3X,s 12
XIs2
Z=90
X26 XI2
X, X,20 and integers
fathomed X,X,20 and integers
Jathomed
The solution of P, is non
the node P is also oi-integer and its Z= 80 and Zy=90. Since, Zy <(Current best Z,
and it has infeasible =90).
constraints for the decision variables. solution in terms of not fulfilling integer
The Problem P, has integer
bound (Z) is updated to the solution. So, itis a fathomed node. Hence the
objective function alue 90. current best lower
COMPILED BY TSEGAY BERHE MSC IN
PRODUCTION ENGINEERINGAND
MANAGEMENT 10
 Now, the Ony unfathomed terminal node is P3. Thefurther branching from this node is show
below:

Max Z=10X +20X,

by tsegay b Subject to:
6X, +8X, s 48

K, X,20 md itegers
X25 X4
P
solution Max Z=10X, +20X, solution Max. Z= 10X +20X,
Scbject to: |Subject to:
X,94 X-83 6X, +SK, s 48
Zu-95 X +3K;S 12 Zy-9333 X+3X,s 12
Z-90 Z=80
, X,20 nd integers K, X;20 mnd mtegers
X23
P P

solution Max Z=10NX, +20X, solution Max Z=10X, +20X,

Ssbject to: |Subject to:
6x, +SX;s48 6x, +SX; s48
Z=90 |X+3X;s12 Z-80 X+3X; S12
Z90 Z=80

X, X20 nd ztegers x,X,20 zmd ntegers

The problems P8 and P9 have integer solution. So, these two nodes are fathomed. But the
objective function value of these nodes are not greater than the current best lower bound of 90.
Hence, the current best lower bound is not updated. Now, all the terminal nodes are fathomed.
The feasible fathomed node with the current best lower bound is Pz. Hence, its solution is treated
as the optimal solution: X=5, X}=2, Z/Optimum) = 90
NOTE: This Problem has alternative optimum solution at Ps with X=3, X=3, Z(Optimum)=90

cOMPILED BY TSEGAY BERHEMSC N PRODUCTION ENGINEERING AND MANAGEMENT 11

 Complete Tree:

tey tett

bet

Soltion P
sbe siho Ssbett aatia

Siet

2lates
uthoned fathoned
Sitinn Solitisn

Bbitt

2lnditegs 2ats
farhmed

cOMPILED BY TSEGAY BERHE MSC IN PRODUCTION ENGINEERINGAND MANAGEMENT 12

 Example 2;
Max Z = 6X, + 8X,
Subject to:
4X, + 5X,s22
5X, + 8X,s 30
X,. X,>0 and integers
b. Integer simplex method (Gomory's cutting plane method)
An Algorithm for solving Pure integer and mixed integer programming
problems have been developed by Ralph E. Gomory
Step I. Relax the integer requirements.
Step II. Solve the resulting LP problem using Simplex Method.
Step III. If all the basic variables have integer values, Optimality of the Integer
programming problem is reached. So, go step 7: otherwise go to step 4.
Step IV. Examine the constraints corresponding to the current optimal solution. For each
Basic Variable with non-integer solution in the current optimal table, find the fractional
part, f, Therefore, b; = [b] +, where [b) is the integer part of bi, and fi is the
fractional part of bi.
Step V. Choose the largest fraction among various fi: ie. Max (fi). Treat the constraint
corresponding to the maximum fraction as the source row
(equation). Based on the source equation, develop an additional constraint (Gomory 's
constraint /fractional cut) as shown:
fi=S;- Summation ((fi) (Non-Basic Variable))
Step VI. Add the fractional cut as the last row in the latest optimal table and
further using dual simplex method, and find the new optimum solution.proceed If the new
optimum1ssolution is integer then go to step 7;
otherwise
Step VII. Print the integer solution [X's and Z- Values]
go to step IV.

Find solution using integer simplex method (Gomory's cutting plane

MAX Z=xl + x2 method)
subject to
3X, + 2X,<5
X,< 20
and X, X, >=0

The optimal table of this linear programming problem is

simplex method. The optimal table obtained at the third iteration of the
looks like as follow:
Iteration-3 C
B CB XB 2 S S2 Min Ratio

cOMPILED BY TSEGAY EERHE MSC IN

PRODUCTION ENGINEERING AND MANAGEMENT 13
 0.5 0
1.5
1

0
2.5

17.5 -1.5
-| -0.5 1

1.5 0.5
Z-2.5

0.5 0 0.5
Zj-Cj

Since all Zi-C20

variables as:
Hence, non-integer optimal solution is arrived with value of
X=0, X;=2.5

Max Z-2.5

Gomory's fractional cut, with the

To obtain the integer valued solution, we proceed to construct
help of X-rowas follows:
2.5=1.5X;tIX;+0.5S;

(2+0.5) = (1+0.5) X,+(l+0) X,t(0+0.5) S,

The fractional cut will become
-0.5=Sgl-0.5X,-0.5SCut-1)
Adding this additional constraint at the bottom of optimal simplex table. The new table so
obtained is:
Iteration-I C 1 0 0

B CBv X X S S Sgl
2.5 1.5 0.5 0 0

S --- 17.5 -1.5 -0.5 1

Sgl -0.5 (-0.5) -0.5 0

Z-2.5 1.5 1 0.5 0 0

Zj-cj 0.5 0.5 0 0

Ratio-Zj-CG/Sgl,j -1‘ -1
and Sgl,j<0

cOMPILED BY TSEGAY BERHE MSCIN PRODUCTION ENGINEERINGAND MANAGEMENr 14

 Minimum negative X¡ is -0.5 and its row index is 3. So, the leaving basis variable is Sgl.
Maximun negative ratio is -l and its column index is 1. So, the entering variable is X,.
The pivot element is -0.5.

Entering =X), Departing =Sgl, Key Element =-0.5

Rold) = -0.5 -05 -0.5 0

Rs(new)=Rold)+ (-0.5) 2

Ri(old) = ---- 2.5 1.5 0.5 0

Rs(new) = -2

1.5xR(new) = 1.5 1.5 1.5 0 -3

Ri(new)=R,(old) - 1.5Rs(new) 0 -I 3

Rz(old) 17.5 -1.5 -0.5

Rs(new) = 1 0 -2

1.5×Rs{new) = 1.5 1.5 1.5 -3

Rz(new)=Rz(old) + 1.5Rs(new) 19 0

Iteration-2 Cj 0
B CBv X |-
X SI S2 Sgl
0 -1 3
S 19 -3
0 0 -2
Z=2 0

Zj-j 0 0

Ratio

COMPILED BY TSEGAY BERHE MSO IN PRODUCTION ENGINEERINGAND

MANAGEMENT
 Since all Zj-Cj20. Hence, integer optimal solution is arrived with value of variables as:
Xj=1,X;=l
Max Z-2

The integer optimal solution found after 1-cuts.

cOMPILED BY TSEGAY BERHE MSCIN PRODUCTION ENGINEERING AND MANAGEMENT 16