Sei sulla pagina 1di 111

WILP-BITS PILANI-BITS-TECHNIP

COLLABORATIVE PROGRAM
MS IN ENGINEERING MANAGEMENT
EMAL ZG612-METHODS AND TECHNIQUES OF
SYSTEM ENGINEERING

P R O F V .M U R A L IDH A R

Applications of Operations ResearchDesign


Implementation
Operations of large, humanly convinced soft systems

Linear programming
Queuing Theory

Integer programming
Inventory control

Simulation

Maintenance Models

Sampling Techniques
Decision Models

Forecasting

Techniques
Network scheduling
Models

Mathematical Models
(1)Operation Research Models (2)Statistical Models

(1)Operation Research Models :


Decision Variables
Objective Function
Constraints
Feasible Solution
Optimal Solution

OR-ORIGIN : World war-I & II


Anti-submarine warfare-Thomas Alva Edison
devised a war game to be used for simulating problems of naval manoeuvre
lpp model for United states economy

OR is a quantitative technique that

deals with many management problems


OR involves the application of scientific methods in
situations where executive requires description,
prediction and comparison
OR is an experimental and applied science devoted
to observe understanding and predicting the
behavior of purposeful man-machine system

APPLICATIONS
AREAS OF
Accounting facilities :Cash flow planning, credit policy planning

of delinquent account strategy


Manufacturing : max or min of objectives of products etc.
Marketing : selections of product mix , productions scheduling
time , advertising allocation etc
Purchasing : How much to order ? Depending on sale or profit etc.
Facility panning : warehouse location, Transportation loading and
un loading, Hospital planning
Finance : Investment analysis , dividend policy etc.
Production : OR is useful in designing,selecting and locating
sizes,scheduling and sequencing the production runs by proper
allocation of machines etc.

TYPES OF OR MODELS
ANALOGUE MODEL (DIAGRAMATIC)
SYMBOLIC MODEL (PHYSICAL MODEL)
DESCRIPTIVE MODEL

METHODOLOGY OF OR MODEL
Identifying the problem
Formulating the problem
Finding out the key areas of the problem
Constructing a mathematical model
Deriving the solution to the model
Testing and updating the model for feasibility of the

solution
Establishing controls over the model and the solution
Implementing the solution obtained

LINEAR PROGRAMING PROBLEM


(LPP)

Optimization of linear function of several variables


Optimization-(maximization or minimization)
Objective function
Decision variables
Constraints
Examples : maximization of profit or production of

any company/Industry
Minimization of expenditure/loss of any
company/Industry

Slack variables and surplus variables


CONSTRAINTS OF LESS THAN OR EQUAL TO TYPE IS CONVERTED
INTO EQUALITY BY INTRODUCING SLACK VARIABLES.
CONSTRAINTS OF GREATER THAN OR EQUAL TO TYPE IS CONVERTED
INTO EQUALITY BY INTRODUCING SURPLUS VARIABLES.
EXAMPLES :

FORMULATION OF AN LPP MODEL


(PRODUCTION-ALLOCATION PROBLEM)
A Manufacturer produces two types of products X and Y.

Each X model requires 4 hours of grinding and 2 hours of


polishing whereas each Y model requires 2 hours of
grinding and 5 hours of polishing. The manufacturer has 2
grinders and 3 polishers. Each grinder works for 40 hours a
week and each poisher works for 60 hours a week. Profit on
X product is Rs3.00 and on Y product it is Rs4.00 .
Whatever is produced in a week is sold in the market. How
should the manufacturer allocate his production capacity to
the two types of products so that he may make optimum
profit in a week?

ANSWER
Max z= 3x+4y
Subject to the constraints

4 x 2 y 80
2 x 5 y 180
with nonnegative restrictio ns x 0, y 0

PRODUCT-MIX PROBLEM
The manager of an oil refinnery has to decide upon

the optimal mix of two possible blending process of


which the inputs and outputs per production run are
as follows :

process

1
2

input

output

Crude-A

Crude-B

Crude-C

Crude-D

5
4

3
5

5
4

8
4

respectively. Market requirements show that at least 100 units of gasoline X and
80units of gasoline y must be produced. The profit per production run from
process 1 and 2 are Rs3 and Rs4 respectively. Formulate the problem as a Lpp
model.

Answer

Max z 3x 4 y subject to the constra int s


5 x 4 y 200
3 x 5 y 150
5 x 4 y 100
8 x 4 y 80
with nonnegative restrictio ns x 0, y 0

METHODS OF OBTAINING OPTIMUM SOLUTIONS


GRAPHICAL METHOD
SIMPLEX METHOD
DUALITY METHOD
DUAL SIMPLEX METHOD
TWO-PHASE METHOD

GRAPHICAL METHOD
APPLIED ONLY WHEN THERE ARE TWO VARIABLES ONLY
CONVERTING INEQUALITY CONSTRAINTS TO EQUALITY

CONSTRAINTS
DETERMINING THE POINTS BY PUTTING EACH
VARIABLES EQUATED TO ZERO
PLOTTING OF GRAPH
IDENTIFYING FEASIBLE REGION
FEASIBLE POINTS TO BE DETERMINED
FINDING THE VALUE OF OBJECTIVE FUNCTION AT
FEASIBLE POINTS
FINDING OPTIMUM SOLUTIONS

Max z= 5x+4y
Subject to the constraints

6 x 4 y 24
x 2y 6
y x 1
y2
with x, y 0

X=3 and y=3/2


OPTIMUM SOLUTION IS GIVEN BY

Minimization Model
1.Find the minimum value of Z=4x+7y
subject to the constraints

x y 6
x 5y 5
x5
with x, y 0

Solution
X=0,Y=1 AND MIN Z=7

on the form. Because of the need to ensure certain nutrient constituents, it is


necessary to buy additional one or two products which we shall call A and B.The
nutrient constituents(vitamins and proteins) in each unit of the production are
given below :

Nutrient
contents in the
products

Nutrient
contents in the
products

Minimum
amount of
nutrient

Nutrient

36

108

12

36

20

10

100

Product A costs Rs20 per unit and Product B costs Rs40 per unit. How much of the products A and
B should be purchased at the lowest possible cost so as to provide the pigs nutrients not less than
that the minimum required as given in the table.

Formulation and Solution :


Minimize Z= 20x+40y

36 x 6 y 108 ( for nutrient 1)


3 x 12 y 36 for nutrient 2
20 x 10 y 100 for nutrient 3

Solution : x=4 and y=2 and min Z = Rs160

x, y 0

3.Find the minimum value of Z=20x+10y


Subject to the constraints

x 2 y 40
4 x 3 y 60
3 x y 30
x, y 0

X=6,Y=12 ; MIN z=240


SOLUTION

Some exceptional cases


An unbounded solution :
While solving lpp, there are situations when the

decision variables are permitted to increase infinitely


without violating the feasibility condition, i.e., no
upper bound. In such a case the objective function
value can also be increased infinitely and hence there
is an unbounded solution.

Example : Consider the lpp :


Maximize Z= 3x+2y

Subject to the constraints

x y 1
x y 3
with x, y 0

The following feasible region (solution space) shows that the


maximum value of Z occurs at a point at infinity and thus the
problem has an unbounded solution
Graph :

No Feasible solution
If it is not possible to find a feasible region that

satisfies all the constraints of the lpp, the given lpp is


said to have feasible solution.This shows that the two
inequalities that form the constraints set are
inconsistent. As there is no set of points that satisfies
all the constraints, there is no feasible solutionto the
problem.

Example : Consider the lpp


Maximize Z =x+y

Subject to the constraints


x y 1
3x y 3
x, y 0

2.Maximize Z=5x+3y
Subject to the constraints

3 x 4 y 12
x 2
x, y 0

Show that the following lpp has an unbounded solution


Minimize Z=6x-2y
Subject to the constraints

2x y 2
x 4,
with x, y 0

Some Important types of solutions of a LPP


SOLUTIONS

DEGENERATE

FEASIBLE SOLUTION

SOLUTION
OPTIMAL SOLUTION
UNBOUNDED
SOLUTION

BASIC FEASIBLE

SOLUTION

SIMPLEX METHOD
STEP-1 : standard form of LPP :
MAX OR MIN (Z)
SUBJECT TO CONSTRAINTS
Ax< or > B
STEP-2 : Check whether the objective function is to

be max or min. If it is min then it can be converted to


max type by using the result
Min(z)=Max(-z)

Continued
STEP-3 :Determine a starting basic feasible solution

by setting decision variables to at zero level and


slack or surplus variables to nonzero level.
STEP-4 : Establish simplex tableau which exhibits
the system of constraints :

Cj

CV

X1

X2

X3

S1

S2

S3

Ratio

S1

9/6

S2

1/3

S3

2/8

Zj

25

-1

21

-3

(ZjCj)

If the problem is maximization type for


optimality solution (Zj-Cj) row must be
>=0
IF THE PROBLEM IS MINIMIZATION TYPE
FOR OPTIMALITY SOLUTION (ZJ-CJ) ROW
MUST BE <=0

Continued
STEP-5 :
Cj row --- The value of cost coefficients of the variables in the objective

function
C column Cost coefficients of non-basic variables in the objective function
CV column- Basic variables column
X1,x2,s1,s1, ---Decision,slack or surplus or artificial variables
RatioRatio column
Zj ---rowMultiplying key column(Incoming variable column) with C
column and adding
(Zj-Cj) rowNet profit
Bold row--- Key row(outgoing variable row) obtained by choosing minimum
element of ratio column
Pivotal element--- Intersection element of key row and key column

Continued
STEP-6 :
To get new simplex table we proceed as follows :
If optimality is not obtained we move to next simplex

table for better solution :


Choose (For Max)most negative value of (Zj-Cj) row
and (For Min)largest positive value of
(Zj-Cj) row. Corresponding column denotes key
column which is Incoming (New Basic
variable)variable column.

Continued
STEP-7 :
Divide each element of key column by the

corresponding element of Cost variable column to


find ratio column.
Choose minimum value of ratio which indicates key
row(Outgoing variable row(leaving the basic column))
Intersection element denote the pivotal element.
STEP-8 : For new simplex table divide key row
element by pivotal element to make it unity(For new
basic variable).

Continued
STEP-9 :The remaining new elements in the simplex

table can be obtained by using the following


formula :
New Value = Old Value --Element correspond ing to key row X
Element correspond ing to key column
Pivotal element

Continued
STEP-10 :
Proceeding as in step-6 till we get optimal or

optimum solution.

Minimization problems
Use of Artificial variables :
In the constraints of the > type , to obtain the initial basic feasible

solution we first put the given lpp in the standard form and then a
nonnegative variable is added to the left side of each equation that
lacks the much needed starting basic variables. This added
variable is called artificial variable.
The artificial variables plays the same role as a slack variable in
providing the initial basic feasible solution . The method will be
valid only if we are able to force these artificial variables to be out
or at zero level when the optimum solution is attained.
In other words, to get back to the original problem artificial
variables must be driven to zero in the final solution; otherwise
the resulting solution may be infeasible.

TWO METHODS FOR THE SOLUTION OF LPP HAVING


ARTIFICIAL VARIABLES :

I. Big-M method or Method of Penalties


II. Two-Phase Method

I .BIG-M Method OR Method of Penalties


STEP-I In this method we assign a very high penalty(say M) to the artificial variables in the

objective function.
STEP-II Write the given L.P.P into its standard form and check whether there exists a starting
basic solution : (a)If there is a ready starting basic feasible solution, move onto step-IV
(b)If there does not exists a ready starting basic feasible solution, move onto step-III .
STEP-III Add artificial variables to the left side of each equation that has no obvious starting
basic variables.Assign a very high penalty (say M) to these variables in the objective function.
STEP-IV Substitute the values of artificial variables from the constraint equations into the
modified objective function.
STEP-V Apply simplex method to the modified L.P.P. Following cases arise at the last
iteration :
(a) At least one artificial variable is present in the basic solution with zero value. In such case
the current optimum basic feasible solution is degenerate.
(b) At least one artificial variable is present in the basic solution with a positive value. In such
a case, the given L.P.P does not posses an optimum basic feasible solution. The given problem
is said to have a pseudo-optimum solution.

Minimize z=4x+y
SUBJECT TO THE CONSTRAINTS

3x y 3
4x 3y 6
x 2y 4
x, y 0

Solution :
Using s1 as surplus variable and s2 as slack variable

with zero cost coefficients in the objective function


and adding artificial variables R1 and R2 with high
penalty value M (MINIMIZING) we get
3x+y+R1=3
4x+3y-s1+R2=6
X+2y+s1=4
With x,y,s1,s2,R1,R20

Continued
Initially we let x=0,y=0,s2=0 (Non-Basic level)
Then we get R1=3,R2=6 and s2=4(Basic level)

Simplex table

Cj

100

100

BV

CV

R1

R2

s1

s2

ratio

R1

R2

-1

6/4

S2

Zj

Zj

900

700

400

100

100

-100

696

399

-100

Zj-Cj

Cj

100

BV

CV

R2

s1

s2

ratio

1/3

R2

5/3

-1

6/5

s2

5/3

9/5

Zj

4/3

Zj

204

504/ 100
3

-100

501/
3

Zj-Cj

Cj

BV

CV

s1

s2

3/5

1/5

6/5

-3/5

s2

Zj

18/5

1/5

1/5

Zj-Cj

ratio

Minimize z=12x+20y
subject to constraints

6 x 8 y 100
7 x 12 y 120
x, y 0

X=15,Y=5/4 AND Min z=205


ANSWER :

3.Max W=2x+y+3z
subject to the constraints

x y 2z 5
2 x 3 y 4 z 12
x, y , z 0

X=3,y=2,z=0
and max w=8
SOLUTION

DUALITY in LINEAR PROGRAMMING


Duality states that for every lpp of max(or min),

there is a related unique problem of min(or max)


based on the same data and the numerical values of
the objective function to the two problems are same.
The original problem is called PRIMAL
PROBLEM while the other is called DUAL
PROBLEM.
Dual problem we define in such a way that it is
consistent with the standard form of the primal.

Examples
I Standard Primal :
Max Z=C1X1+C2X2+C3X3+..CnXn
Subject to the constraints :
A11X1+a12X2+a1nXn=bi, (i=1,2,3,m)
Dual :
Min Z=b1y1+b2y2+b3y3+bmym
Subject to the constraints :
A1jy1+a2jy2+amjymCj, (j=1,2,3,n)

yi unrestricted

(i=1,2,3.m)

I Standard Primal :
Min Z=C1X1+C2X2+C3X3+..CnXn
Subject to the constraints :
ai1X1+ai2X2+ain Xn=bi, (i=1,2,3,n)

Xj0 ; (j=1,2,3.m)

Dual :
Min Z=b1y1+b2y2+b3y3+bmym
Subject to the constraints :
A1jy1+a2jy2+amjynCj, (j=1,2,3,n)

yi unrestricted

(i=1,2,3.m)

OPTIMAL DUAL SOLUTION :The following set of rules govern the


derivation of the optimum dual solution
Rule-1 :If the primal(dual) variable corresponds to a slack

and/or surplus variable in the dual(primal) problem, its


optimum value is directly read off from the last row of the
optimum dual(primal) simplex table, as the value corresponding
to this slack and/or surplus variable.
Rule-2 :If the primal(dual) variable corresponds to artificial
starting variable in the dual(primal) problem, its optimum value
is directly read off from the last value row of the optimum
dual(primal) simplex table, as the value corresponding to this
artificial variable, after deleting the constant M, the penalty cost.
Rule-3 : If the primal(dual) problem is unbounded , then the
dual(primal) problem does not have any feasible solution.

PROBLEMS
1.Formulate the dual of the following linear

programming problem :
Maximize z 5 x1 3 x2
Subject to the constra int s
3 x1 5 x2 15,
5 x1 2 x2 10,
x1 , x2 0

Solution :
Introducing slack variables

s1 0, s2 0

The reformulated linear programming problem is

Maximize z 5 x1 3x2 0.s1 0.s2


Subject to the constra int s
3 x1 5 x2 s1 0.s2 15,
5 x1 2 x2 0.s1 s2 10,
x1 , x2 , s1 , s2 0

Corresponding Dual is given by

Minimize w 15 y1 10 y2
Subject to the constra int s
3 y1 5 y2 5,
5 y1 2 y2 3,
y1 , y2 0
y1 and y2 unrestrict ed (redundant )

2.Find the dual of the following maximization problem :

Maximize z 2 5 x1 35 x2
Subject to the constra int s
2 x1 3x2 15,
x1 4 x2 20,
3 x1 x2 12
where x1 , x2 0

SOLUTION :

Maximize z 25 x1 35 x2 0.s1 0.s2 0.s3


Subject to the constra int s
2 x1 3x2 s1 0.s2 0.s3 15,
x1 4 x2 0.s1 s2 0.s3 20,
x1 , x2 , s1 , s2 , s3 0

The corresponding duality is given by


Minimize w 15 y1 20 y2 12 y3
Subject to the constra int s
2 y1 y2 3 y3 25,
3 y1 4 y2 y3 35,
y1 , y2 , y3 0
y1 and y2 unrestrict ed (redundant )

3.Write the dual of the following linear programming problem :

Minimize w 5 x1 6 x2 4 x3
Subject to the constra int s
3 x1 4 x2 6 x3 9,
x1 3x2 2 x3 5,
7 x1 2 x2 x3 10
x1 2 x2 4 x3 4
2 x1 5 x2 3x3 3
x1 , x2 , x3 0

Reformulating in the standard form we get

Minimize z 5 x1 6 x2 4 x3 0.s1 0.s2 0.s3 0.s4 0.s5


Subject to the constra int s
3x1 4 x2 6 x3 s1 0.s2 0.s3 0.s4 0.s5 9,
x1 3x2 2 x3 0.s1 s2 0.s3 0.s4 0.s5 5,
7 x1 2 x2 x3 0.s1 0.s2 s3 0.s4 0.s5 10,
x1 2 x2 4 x3 0.s1 0.s2 0.s3 s4 0.s5 4,
2 x1 5 x2 3 x3 0.s1 0.s2 0.s3 0.s4 s5 3,
with x1 , x2 , x3, s1 , s2 , s3 , s4 , s5 0

The corresponding duality is given by

Maximize w 9 y1 5 y2 10 y3 4 y4 3 y5
Subject to the constra int s
3 y1 y2 7 y3 y4 y5 5,
4 y1 3 y2 2 y3 2 y4 5 y5 6,
6 y1 2 y2 y3 4 y4 3 y5 4,
y1 , y2 , y3 , y4 , y5 0
y1 , y2 , y3 , y4 and y5 unrestrict ed (redundant )

4.One unit of product A contributes Rs7 and requires 3 units of raw materials and 2 hours of labour. One unit of product B contributes Rs 5 and requires one unit of raw material and one unit of labour. Availability of raw material of present is 48 units and there are 40 hours of labour.

(i)Formulate it as a linear programming problem


(ii)Write its dual
(iii)Solve the dual with simplex method and find the

optimal product mix and the shadow prices of the


raw material and labour.

Solution :

Maximize z 7 x1 5 x2
Subject to the constra int s
3 x1 x2 48,
2 x1 x2 40,
x1 , x2 0

Dual problem
Minimize w 48 y1 40 y2
Subject to the constra int s
3 y1 2 y2 7,
y1 y2 5,
y1 , y2 0

SOLUTION
OPTIMUM SOLUTION OF DUAL VARIABLES ARE

y1 0 and y2 5

ORIGINAL SOLUTION ARE


x1 0 and x2 40 and
min( w) max( z ) 200

Integer programming-An LPP wiTH ADDITIONAL REQUIREMENTS


THAT THE VARIABLES CAN TAKE ON ONLY INTEGER VALUES IS
CALLED iNteger programming

Max or Min Z=C1X1+C2X2+C3X3+..CnXn


Subject to the constraints :
ai1X1+ai2X2+ain Xn=bi, (i=1,2,3,m)
and

Xj0 ; (j=1,2,3.n)
Where Xj are integer valued for j=1,2,3.p (pn)
If p=n , the problem is called Pure Integer
programming problem , otherwise it is called Mixed
Integer programming problem.
Also if all the variables of an integer programming problem
are either 1 0r 0, the problem is called Zero-one
programming problem.

CAPITAL BUDGETING PROBLEM


Five projects are being evaluated over a 3-year

planning horizon .The following table gives the


expected returns for each project and the associated
Expenditures(milli
yearly expenditures.
ons $) per year

Project

Returns(million$)

20

10

40

20

15

10

30

Available
funds(milli
on$)

25

25

25

solution

Maximize z 20 x1 40 x2 20 x3 15 x4 30 x5
Subject to the constra int s
5 x1 4 x2 3x3 7 x4 8 x5 25,
x1 7 x2 9 x3 4 x4 6 x5 25,
8 x1 10 x2 2 x3 x4 10 x5 25,
x1 , x2 , x3 , x4 , x5 (0,1)

The Optimum integer solution is

x1 x 2 x3 x 4 1, x5 0
and max z 95 ( million $)

THE SOLUTION SHOWS THAT ALL BUT


PROJECT-5 MUST BE SELECTED

0 x1 , x2 , x3 , x4 , x5 1

Solution is

x1 0.5789 , x 2 x3 x 4 1, x5 0.7368
and max z 108.68 ( million $)

INTEGER PROGRAMMING ALGORITHMS


Step-1 Relax the solution space of the ILP by deleting

the integer restrictions on all integer variables and


replacing any binary variable y with the continuous
range 0y1.
The result of the relaxation is a regular LP.
Step-2 Solve the LP and identify its continuous
optimum.
Step-3 Starting from the continuous point , add
special constraints that iteratively modify the LP
solution space satisfying the integer requirements.

For generating special constraints there are two general methods :

1.Cutting Plane Method


2.Branch and Bound method

1.Cutting Plane Method


Step-1Put the given LPP into its standard form and determine

the optimum solution by using simplex methods ignoring


integer value restrictions.
Step-2 Test for integerality of the optimum solution :
(i) If the optimum solutionGadmits
f f all
x
integer values, an
optimum basic feasible solution
where 0 fis
1 attained.
(ii) If the optimum solution does not include all integer values
then go to the step-3
Step-3 Choose a row corresponding to the basic variable which
has largest fractional cut , say, fk
And generate the constraint
in the form
n
n

k0

j 0

kj

kj

G1 f k 0 f kj x j
j 0

where 0 f kj 1 and 0 f k 0 1

Continued .
Step-4 Add the constraint to the optimum simplex

table obtained in step-1. Apply dual simplex method


to find an improved optimum solution.
Step-5 Go to step-2 and repeat the procedure until
an optimum basic feasible all integer solution is
obtained.

Solve the integer programming problem

Maximize z 60 x1 40 x2
Subject to the constra int s
2 x1 x2 60,
x1 25
x2 35,
x1 , x2 0 and are int egers

Standard form

Maximize z 60 x1 40 x2 0.s1 0.s2 0.s3


Subject to the constra int s
2 x1 x2 s1 0.s2 0.s3 60,
x1 s2 25,
x2 s3 35
x1 , x2 , s1 , s2 , s3 0

TRANSPORTATION MODEL
Definition :There are m sources and n destinations which are

represented by a node.
ith source and jth destination are linked by the arc (i,j). The
transporation cost from ith source to jth destination is represented by

cij

The amount shipped by

xij

The amount of supply at source by

The amount of demand at destination j by

ai
bj

The objective of the model is to determine the unknown

xij
that will minimize
the total transportation cost while satisfying all the

supply and demand restrictions.

GENERAL TRANSPORTATION MODEL


A

SUPPLY

X11 (2)

X12 (35)

x13

x14

b1

x21

X22 (23)

x23

x24

b2

X31 (18)

x32

X33 (42)

x34

b3

DEMAND

a1

a2

a3

a4

Grand
Total

PRODUCTION INVENTORY CONTROL


A company manufactures backpacks for serious hikers. The

demand for its product occurs from March to June of each year.
The company estimates the demand for the four months to be
100,200,180,and 300 units. The company uses part time labour
to manufacture the backpacks and as such its production
capacity varies monthly. It is estimated that company can
produce 50,180,280, and 270 units for March to June,
respectively.Because the production capacity and demand for
different months do not match, a current months demand may
be satisfied in one of the three ways :
1. Current months production
2. Surplus production in earlier months
3. Surplus production in later months

Continued
In the first case, the production cost per backpack is

$40.00. The second case incurs an additional


holding cost of $50 per backpack per month. In the
third case, an additional penalty cost of $2.00 is
incurred per backpack per month delay. The
company wishes to determine the optimal
production schedule for the four months.

SOLUTION :TRANSPORTATION MODEL


1

CAPACITY

40

40.50

41.00

41.50

50

42.00

40.00

40.50 41.00

180

44.00

42.00

40.00 40.50

280

46.00

44.00

42.00 40.00

270

DEMAN
D

100

200

180

300

The unit transportation cost from period I to period j is


computed as :

Cij = production cost in i, i=j

production cost in I + holding cost


from i to j , i<j
production cost in i +penalty cost
from i to j , i>j

TRANSPORTATION ALGORITHM
I NORTH WEST CORNER RULE METHOD
II LEAST COST METHOD
III VOGELS APPROXIMATION METHOD

Transportation algorithm-vogels approximation method (vam)


VAM method is an improved version of the least cost method that generally produces better

starting solution :
STEP-1
For each row(column), determine a penalty measure by subtracting the smallest unit cost
element in the row(column) from the next smallest unit cost element in the same row
(column)
STEP-2
Identify the row or column with the largest penalty. Break ties arbitarily. Allocate as much
as possible to the variable with the least unit cost in the selected row or column. Adjust the
supply and demand , and cross out the satisfied row or column. If a row and a column are
satisfied simultaneously, only one of the two is crossed out, and the remaining row(column)
is assigned zero supply(demand)
STEP-3
(a) If exactly one row or one column with zero supply or deman remains uncrossed out,
stop.
(b)If one row(column) with positive supply(demand) remains uncrossed out, determine the
basic variables in the row(column) by the least cost method.
(c) If all the uncrossed out rows and columns have (remaining) zero supply and demand,
determine the zero basic vaiables by the least cost method. Stop.
(d) Otherwise go to step-1

Problems-1
Warehous
e

Stores

Availabilit
y

II

III

IV

34

15

12

-1

19

25

17

17

80

Requirement 21

TRANSPORTATION COST IS GIVEN BY


6X5+6X1+17X3+5X3+15X3+12X3+19X-1=rs 164

Wareh
ouse

Stores

Availab
ility

II

III

IV

5 (6)

1 (6)

3 (17)

3 (5)

34

3
(15)

15

3 (12)

12

-1 (19)

19

Require
ment

21

25

17

17

80

INITIAL BASIC FEASIBLE SOLUTION USING VAM METHOD


Warehouse

stor
es

Avail
abilit
y

Row
penalties

II

III

IV

34

2 2 0 0

15

0 0 1

12

1 1

-1

19

3 - -

Requirement

21

25

17

17

80

Column
penalties

Problem-1 ANSWER : COST=rs430


DESTINA
TION
ORIGIN

D1

D2

O1

13

15

O2

O3
Demand

Supply
D3

17

16 (11)

12

11

2 (12)

16

10 (14)

20 (2)

14

23

(6)

45

Problem-2 ANSWER : COST=rs370


DESTI
NATIO
N

Supply

ORIGI
N

D1

D2

D3

D4

O1

5 (10)

3 (20)

O2

4 (20) 5 (20)

4 (10) 50

O3

2 (20)

20

Deman
d

30

40

20

10

100

30

Modi method (optimality solution)


Modi method helps to determine better transportation schedule by

evaluating the non-basic cells. The Modi method selects a particular


nonbasic cell that will yield maximum improvement through a set of
index number for the row and column cells. The movement of
maximum permissible units are then made for the non-basic cell. In a
Similar manner, revised index numbers indicate the next best nonbasic cell. This is based on the following
principle :
For a basic feasible solution of dual variables ui (i=1,2,3m)
and vj(j=1,2,3n) corresponding to capacity constraints i
and requirement constraints j satisfying
Cij = ui+vj for each basic cell(occupied cell) then the index
number (opportunity cost) Dij corresponding to each nonbasic cell(I,j) is given by Dij = ui+vj-Cij

Problem-4 : find the minimum transportation cost using


modi method
TO

FROM

Require
ment

II

III

IV

Availability

15

10

17

18

16

13

12

13

12

17

20

11

15

Answer : Initial transportation cost rs 320


optimum cost rs174

ASSIGNMENT MODELS : The assignment problem is a special case of the transportation problem in which the objective is to assign a number of origins to the equal number of destinations at a minimum cost (or maximum profit). The assignment is to be made on a one-one basis.

Mathematical Formulation :
Min Z= CijXij
Subject to the constraints
Xij =1 and Xij =1 ; xij = 0 or 1
For all i=1,2,3,4, and j=1,2,3,4
Where Cij is the cost associated with assigning ith

resource to jth activity.

REDUCTION THEOREM
TH-1 : In an assignment problem , if we add or

subtract a constant to every element of any row(or


any column) of the cost matrix [Cij], then an
assignment that minimizes the total cost on one
matrix also minimizes the total cost on the other
matrix.
TH-2 :If Cij0 such that the minimum Cij = 0,
then Xij provides an optimum assignment.

HUNGARIAN METHOD (Hungarian mathematician D.KONIG)


Step-1
Determine the cost table from the given problem
(i) If the number of sources is equal to the number of destinations go

to step-3.
(ii) If the number of sources is not equal to the number of destinations
go to step-2.
Step-2
Add a dummy source or dummy destination, so that the cost table
becomes a square matrix. The cost entries of dummy
source/destinations are always zero.
Step-3
Locate the smallest element in each row of the given cost matrix and
then the same from each element of that row

Continued .
Step-4
In the reduced matrix obtained in step-3, locate the smallest element of each

column and then subtract the same from each element of that column. Each
column and row now have at least one zero.
Step-5
In the modified matrix obtained in step-4, search for an optimal assignment as
follows :
(a) Examine the rows succesively until a row with single zero is found. En
rectangle this zer0 and cross off all other zeros in its column. Continue in this
manner untill all the rows have been taken care of.
(b) Repeat the procedure for each column of the reduced matrix.
(c) If a row and/ or column has two or more zeros and one cannot be chosen by
inspection then assign arbitary any one of these zeros and cross off all other zeros
of that row/column.
(d) Repeat the above process until the chain of assigning rectangle or cross ends.

Continued .
Step-6
If the number of assignments are equal to n an optimum solution is reached.
If the number of assignments are less than n, go to next step.
Step-7
Draw the minimum number of horizontal and/ vertical lines to cover all the zeros of the reduced matrix as

follows :
(a)Mark () rows that do not have any assigned zero.
(b) Mark() columns that have zeros in the marked rows.
(c) Mark() rows that have assigned zeros in the marked columns
(d) Repeat above until the chain of marking is completed.
(e) Draw lines through all the unmarked rows and marked columns. This gives the minimum number of
lines.
Step-8
Devel0p the new revised cost matrix as follows :
(a) Find the smallest element of the reduced matrix not covered by any of the lines.
(b) Subtract this element from all the uncovered elements and add the same to all the elements lying at the
intersection of any two lines.
Step-9
Go to step-6 and repeat the procedure until an optimum solution is attained.

PROBLEMS :
Answer : 1-I , 2-III , 3-II , 4-IV ;
Minimum Time = 42 hours

1.A department has four subordinates and four tasks

are to be performed. The subordinates differ in


efficiency and their tasks differ in their intrinsic
difficulties. The estimate of time (in hour) each man
would take to perform each task is given by
Task

Subordinate

II

III

IV

26

17

11

13

28

26

38

19

18

15

19

26

24

10

2.Shortest-Route problem

Find the shortest route between nodes 1 and

7 of the network in the following diagram by


formulating the problem as a transshipment
model. The distance between the different
nodes are shown in the nettwork (Hint :
Assume that node-1 has a net supply of 1 unit
and node-7 has a net demand also of one
unit.)

Network diagram

TYPES OF ASSIGNMENT MODELS


1.Maximum assignment model
2.Restrictions on assignment
3.Prohibited assignment model
4.Travelling salesman problem

Potrebbero piacerti anche