Professional Documents
Culture Documents
announced
Quizzes ** 20 OB
Historically, the first term for optimization was “linear programming”, which
was due to George B. Dantzig, although much of the theory had been
introduced by Leonid Kantorovich
ü Interdisciplinary approach
ü Scientific approach
ü Holistic approach
ü Objective oriented approach
ü Judgement phase
ü Research phase
ü Action phase
§ Excel 97
Advantages
Ø A model describes relationships among various variables more effectively than
verbal communication.
Ø Models allows experiment on complex system that is not possible in real life.
Ø It can simplify investigation.
Ø It can identify bottle neck of the process.
Classification
Ø Deterministic
Ø Probabilistic
Ø Static
Ø Dynamic
Ø Analytical model
Ø Simulation model
Dr. Manoj Kumar Pandey
Models in operations research
Limitations:
Ø In real life problems, decision variables, or objective function
are not linearly related.
Ø Rounding off error cannot be avoided that leads to error in
optimal solution
Ø Effect of time and uncertainty is neglected.
Ø LPP deals with single objective.
Ø Parameters are considered constant.
Dr. Manoj Kumar Pandey
Production example
B 20 1.7
C 45 2.5
The company has a daily order commitment for 20 units of products A and a total of 15
units of products B and C. Formulate the LP model.
Table-1
Day No. of employees required
1 -Monday 10
2 -Tuesday 6
3 -Wednesday 8
4 -Thursday 12
5 –Friday 7
6 -Saturday 9
7 -Sunday 4
An institute has to organize its annual cultural festival continuously for next five
days. There is an arrangement of dinner for every invited team. The requirement of
napkins during these five days is:
Days 1 2 3 4 5
Accordingly, a caterer has been requested to supply the cloth napkins according to
the schedule and no napkins in the beginning. After the festival is over caterer has no
use of napkins. A new napkin cost 50 Rs. The washing charges for a used napkin is
15 Rs by ordinary service and 25 Rs, if express service is used. A napkin given by
ordinary service is returned third day, while under express service it is return next
day. How the caterer should meet the requirement of the festival organizers so that
the total cost is minimized.
Dr. Manoj Kumar Pandey
Dr. Manoj Kumar Pandey
Dr. Manoj Kumar Pandey
Engineering Optimization ( ME F320)
Department of Mechanical Engineering
BITS Pilani K. K. Birla Goa campus
Instructor in charge: Dr. Biswajit Das
Email id: biswajitd@goa.bits-pilani.ac.in
A company has two plants, each of which produces and supplies two products: A and
B. The plants can each work up to 16 hours a day. In plant 1, it takes three hours to
prepare and pack 1000 gallons of A and one hour to prepare and pack one quintal of
B. In plant 2, it takes two hours to prepare and pack 1000 gallons of A and 1.5 hours
to prepare and pack a quintal of B. In plant 1, it costs Rs 15000 to prepare and pack
1000 gallons of A and Rs 28000 to prepare and pack a quintal of B, whereas in plant
2 these costs are Rs 18000 and Rs 26000, respectively. The company is obliged to
produce daily at least 10 thousand gallons of A and 8 quintals of B. Formulate this
problem as an LP model to find out as to how the company should organize its
production so that the required amounts of the two products be obtained at minimum
cost.
1 5 150
2 7 200
3 9 300
Dist centre
Plant 1 2 3 Supply
1 2 3 11 6
2 1 9 6 9
Demand 7 5 3
Dist centre
Plant 1 2 3 Supply
1 2 3 11 6
2 1 9 6 9
Demand 7 5 3
A company produces two types of hats. Type I requires twice as much as labour time
as type II alone. If all labour time is dedicated to type II alone, the company can
produce a total of 400 hats of type II per day. Respective market limits for two types
are 150 and 200 per day. The profit is Rs. 8 for type I hat and Rs. 5 per type II hat.
Find the optimum no. of hats to be produced so that company's profit can be
maximized.
Graphical method:
(i) Extreme point solution method
(ii) Iso-profit (cost) function line method
Solution: Any value of the decision variables x j (j=1,2,..,n) that satisfy all the
constraints of an LP problem is called as a of that LP problem.
Feasible solution: The set of values of the decision variables xj (j=1,2,..,n) that
satisfy all the constraints and non negativity conditions simultaneously of an LP
problem is called as a of that LP problem.
Infeasible solution: The set of values of the decision variables xj (j=1,2,..,n) that
does not satisfy all the constraints and non negativity conditions simultaneously of
an LP problem is called as a of that LP problem.
(a) Replace the inequality sign with equality sign for each constraint.
(b) Draw straight lines on the graph paper and take decision for area of feasible
solutions according to the inequality sign of the constraint.
(c) Shade the common portion of the graph that satisfies all the constraints drawn.
(d) The final shaded area called feasible region (or solution space) of the given LP
problem. Any point inside the shaded region is called feasible solution and these
solutions satisfy all the constraints.
(a) Determine the coordinates of each extreme point of the feasible solution space.
(b) Compute and compare the values of the objective function at each extreme point.
(c) Identify the extreme point that gives optimal (max or min) value of objective
function.
x1 , x2 ≥ 0
x1 , x2 ≥ 0
x1 , x2 ≥ 0
Unbounded solution
occurs when particular decision variables and value of
objective function are permitted to increase infinitely without violating feasibility
condition.
Unbounded solution
Unbounded solution
x1 , x2 ≥ 0
Infeasible solution
arises there is no solution that satisfies all the constraints
simultaneously.
Infeasible solution
X 25% 75%
Y 10% 20%
The price of A’s scrap is Rs 200 per quintal and that of B is Rs 400 per quintal.
The firm wants to determine the quantities that it should buy from the two
suppliers so that the total cost is minimized.
Dr. Manoj Kumar Pandey
Dr. Manoj Kumar Pandey
Dr. Manoj Kumar Pandey
Dr. Manoj Kumar Pandey
Practice problem
The manager of oil refinery must decide on the optimal mix of two possible
blending processes of which the inputs and outputs per production run are as
follows:
2 4 5 4 4
The maximum amounts available of crudes A and B are 200 units and 150 units,
respectively. Market requirements show that at least 100 units of gasoline X and
80 units of gasoline Y must be produced. The profits per production run for
process 1 and process 2 are Rs. 300 and Rs. 400, respectively. Formulate this
problem as an LP model and solve it using graphical method to determine the
production run for process 1 and Dr.
2.Manoj Kumar Pandey
Algebraic method
If the RHS of the equation is negative then multiply both sides of the
equation by -1.
Example 1:
x1 , x2 ≥ 0
An iterative Initialization
procedure (Find initial BFS)
Move to a better
adjacent BFS
• Equivalent to:
Let aij be the coefficient of the entering variable xj in the ith row.
bi
, aij 0 is Minimum
aij
subject to
x1 ≤4
2x2 ≤ 12
3x1+ 2x2 ≤18
x1, x2 ≥ 0
Z - 3x1 - 5x2 =0
x1 +s1 =4
2x2 +s2 = 12
3x1+ 2x2 +s3 = 18
s1
s2
s3
Basic
Z x1 x2 s1 s2 s3 Solution
variable
Z 1 -3 -5 0 0 0 0
s1 0 1 0 1 0 0 4
s2 0 0 2 0 1 0 12
s3 0 3 2 0 0 1 18
s2 0 0 2 0 1 0 12 12/2 =6
18/2 =9
s3 0 3 2 0 0 1 18
Pivot row
New pivot row= Current pivot row/ pivot element
Min.
Basic Ratio
Z x1 x2 s1 s2 s3 Solution
variable
Z 1 -3 0 0 5/2 0 30 No Ratio
s1 0 1 0 1 0 0 4 4/1 = 4
x2 0 0 1 0 1/2 0 6 -------
s3 0 3 0 0 -1 1 6 6/3 =2
Z 1 0 0 0 3/2 1 36
s1 0 0 0 1 1/3 -1/3 2
x2 0 0 1 0 1/2 0 6
x1 0 1 0 0 -1/3 1/3 2
Feasible (4, 3)
Region
(0, 0) (4, 0)
0 s2 800 1 2 0 -1 0 1 800/2=800
0 s3 400 1 0 0 0 1 0 400/1=400
5
Manoj Kumar Pandey, OPTIMIZATION 6
Manoj Kumar Pandey, OPTIMIZATION 7
Iteration 2
0 s2
4 x1
0 s4
8
Manoj Kumar Pandey, OPTIMIZATION 9
Iteration 3
0 s2
4 x1
0 s4
10
Manoj Kumar Pandey, OPTIMIZATION 11
Iteration 4
0 s3
4 x1
0 s4
12
Manoj Kumar Pandey, OPTIMIZATION 13
Engineering Optimization (ME F320)
Department of Mechanical Engineering
BITS Pilani K. K. Birla Goa campus
Instructor in charge: Dr. Biswajit Das
• Big M Method
2
If in a starting simplex tableau, we don’t have an
identity sub matrix (i.e. an obvious starting BFS), then
we introduce artificial variables to have a starting BFS.
3
Big M Method
4
Suppose, ith constraint equation doesn't have a slack
variable. i.e.
ith constraint in the original LPP is either ≥ or = type .
5
Artificial variable objective coefficient
= - M (Sum of artificial variables) in a maximization
problem,
= M (sum of the artificial variables ) in a minimization
problem
where M is a very large positive number.
Mathematically M
6
This problem is solved in class:
x1, x2 ≥ 0
M A2 60 1 2 0 -1 0 1 60/2=30
9
Manoj Kumar Pandey, OPTIMIZATION 10
Manoj Kumar Pandey, OPTIMIZATION 11
Iteration 2
12
Manoj Kumar Pandey, OPTIMIZATION 13
Iteration 3
14
Manoj Kumar Pandey, OPTIMIZATION 15
Engineering Optimization (ME F320)
Department of Mechanical Engineering
BITS Pilani K. K. Birla Goa campus
Instructor in charge: Dr. Biswajit Das
s1
A1
A2
Iteration 2
s1
x2
A2
Iteration 3
s1
x2
x3
Engineering Optimization (ME F320)
Department of Mechanical Engineering
BITS Pilani K. K. Birla Goa campus
Instructor in charge: Dr. Biswajit Das
Subject to
2x1 + 5x2 + x3 = 12
3x1 + 4x2 = 11
-M A1 11 3 -3 4 0 1 11/4
4
Iteration 2
5
Iteration 3
6
Iteration 4
7
Engineering Optimization (ME F320)
Department of Mechanical Engineering
BITS Pilani K. K. Birla Goa campus
Instructor in charge: Dr. Biswajit Das
Ø Degeneracy
Ø Unbounded Solutions
Ø Infeasible Solutions
ALTERNATIVE OPTIMUM SOLUTION
Example
Cj 6 4 0 0 0
Cj 6 4 0 0 0
x2
Example 1 Constraint x1 + 4x2 8 is
redundant as it is not essential
Max z = 3x1 + 9x2 for finding the optimal solution
Subject to
x1 + 4x2 8
x1 + 2x2 4
x1 + 4x2 8
x1, x2≥ 0
x1
x1 + 2x2 4
Cj 3 9 0 0
0 s1 8 1 4 1 0 8/4=2
0 s2 4 1 2 0 1 4/2=2
Rules for selection
s1 4 1 0
s1 2 0 1
Rules for selection
s1 4 ¼=1/4 0/4=0
s2 2 0/2=0 ½=1/2
Iteration 2
Cj 3 9 0 0
0 s1 0 -1 0 1 -2
9 x2 2 1/2 1 0 1/2
Degeneracy
If there is a tie between two decision variables then selection
can be made arbitrarily.
If there is a tie between two slack (or surplus variable), then
selection can be made arbitrarily.
How to remove degeneracy
Cj 3 5 0 0 0 -M
0 s1 6 1 -2 1 0 0 0 ---
0 s2 10 1 0 0 1 0 0 ----
-M A1 1 0 1 0 0 -1 1 1
Iteration 2
Cj 3 5 0 0 0 -M
0 s1 8 1 0 1 0 -2 2
0 s2 10 1 0 0 1 0 0
5 x2 1 0 1 0 0 -1 1
INFEASIBLE SOLUTION
Infeasible solution
This situation can never occur when all the constraints are of the
type ().
Cj 6 4 0 0 -M
0 s1 5 1 1 1 0 0 5/1
-M A1 8 0 1 0 -1 1 8/1
Iteration 2
Cj 6 4 0 0 -M
4 x2 5 1 1 1 0 0
-M A1 3 -1 0 -1 -1 1
Engineering Optimization (ME F344 and ME F320)
Department of Mechanical Engineering
BITS Pilani K. K. Birla Goa campus
Instructor in charge: Dr. Biswajit Das
4
Phase - II
5
Example
Min z = x1 + x2
Subject to
2x1 + x2 ≥ 4
x1 + 7x2 ≥ 7
x1, x2≥ 0
Max z* = -x1 - x2
Subject to
2x1 + x2 –s1 +A1 = 4
x1 +7x2 –s2 +A2 = 7
Max z* = -A1 - A2
Subject to
2x1 + x2 –s1 +A1 = 4
x1 +7x2 –s2 +A2 = 7
Cj 0 0 0 0 -1 -1
-1 A1 4 2 1 -1 0 1 0
-1 A2 7 1 7 0 -1 0 1
Iteration 1
Cj 0 0 0 0 -1 -1
Cj
Cj 0 0 0 -1 -1
-1 A1 2 -2 1 3 1 0
-1 A2 1 2 3 4 0 1
Iteration 1
Cj 0 0 0 -1
17
Engineering Optimization (ME F344 and ME F320)
Department of Mechanical Engineering
BITS Pilani K. K. Birla Goa campus
Instructor in charge: Dr. Biswajit Das
1 c z 0
0 A
X b
Let B is a m x m non-singular sub-matrix of the
coefficient matrix A.
Let XB be the corresponding set of basic variables
with cB as its associated objective vector. The
corresponding solution (and the objective function
value ) may be computed as follows:
1 -cB z 0
X
0 B B b
1
z 1 CB 0
X 0 b
B B
z 1 CB B 0
1
X 1
B 0 B b
z C B B b C B X B
1
X 1 B 1 b
B B b
The general simplex tableau:
1 cB B 1 c z 1 cB B 0
1 1
1 1
0 B 0 A X 0 B b
1 cB B A - C z cB B b
1 1
1 1
0 B A X B b
Thus in the new simplex tableau the z column
will be 1
0
.
.
0
Basic z xj Solution
1 1
z 1 cB B Aj c j cB B b
1 1
XB 0 B Aj B b
Using the notations
a1 1 a1 2 . . a1 n
a a 22 . . a 2 n
21
A .
.
a m 1 am 2 . . a m n
Using the notations
M a x. z C X
S u b je c t to
AX b
X 0
Where
x1 b1
x b
2 2
C c1 c2 . . cn X . , b .
,
. .
xn bm
a11 a12 . . a1 n
a a 22 . . a2 n
21
A .
.
a m 1 am 2 . . a mn
Dual Problem:
T
Min. w b Y
Subject to
T T
A Y C
Y Unrestricted in sign
T
where Y ( y1 y 2 y m )
Write the dual of the LP
Maximize z x1 x2
subject to
2 x1 x2 5
3x1 x2 6
x1 , x2 unrestricted in sign
Engineering Optimization (ME F320)
Department of Mechanical Engineering
BITS Pilani K. K. Birla Goa campus
Instructor in charge: Dr. Biswajit Das
z j c j
, a ij 0
a ij
x2 0 1/2 1 0 -1/2 5
This is the optimal tableau.
M A1 4 -1 -1 -1 0 1 0 ----
M A2 1 -1 1 0 -1 0 1 2
Iteration 2
Cj -3 -2 0 0 M M
M A1 6 -2 0 -1 -1 1 1
-2 y2 2 -1 1 0 -1 0 1
Final table or optimal table
Cj 1 -3 2 0 0 0