Application, History & Basic Formulation Application and Model Building

Outline:
Lecture-1
• Necessity of Optimization and Management Science
DMOT • Resources are scanty and some even can’t regenerated within small span of time
Modelling with Optimization Techniques: Ex: Kerala Flood, War, Production Planning
History, Applications, Importance & Beyond (Analytics) • Model Building Approach with real-time situations
• Sensitivity Analysis
• Interpretation
• Analytics

https://www.flightradar24.com/12.74,82.69/5#

| DMOT | Introduction | Goutam Sutar

DMOT | | Goutam Sutar

1 2

Application and Model Building Application and Model Building

What RAF planned
• Estimated the number of bombers likely to be lost to enemy night fighters and flak,
and how many would be lost through collisions.
• Minimizing the former demanded a densely packed stream, as the controllers of a
night fighter flying a defensive 'box' could only direct a maximum of six potential
Freya radar
interceptions per hour, and the flak gunners could not concentrate on all the available
targets at once.

bomber stream
Kammhuber Line Map
Source: https://en.wikipedia.org/wiki/Operations_research Source:
Search lights, 1944 https://web.archive.org/web/20111105053432/http://www.raf.mod.uk/bombercommand/thousands.h
tml
DMOT | | Goutam Sutar DMOT | | Goutam Sutar

3 4

Application and Model Building Application and Model Building

Also, many believe this helped RAF

History:
• Enigma Machine
• Alan Turing • Application and requirement during WW-II
• Kantorovich (1939) to
George B Dantzig (1947)
•The term “linear” and “Programming”
•If linear, is there any other programming exists? What
are they?
• The RAND Corporation.
•Definition
•Pre-requisites

DMOT | | Goutam Sutar DMOT | | Goutam Sutar

5 6

1
 Application and Model Building Application and Model Building
Major Application Areas: Source :wiki
Some of the fields that have considerable overlap with Operations Research and Management Science include:

Applications: •Business Analytics

•Data mining
•Decision analysis
•Engineering
• The Importance of Linear Programming or Optimization •Financial engineering
– Many real-world problems lend themselves to linear •Forecasting
•Game theory
programming modeling. •Graph theory
– Many real-world problems can be approximated by linear models. •Industrial engineering
– There are well-known successful applications in: •Logistics
•Mathematical modeling
• Manufacturing •Mathematical optimization
• Marketing •Probability and statistics
•Project management
• Finance (investment) •Policy analysis
• Advertising •Simulation
•Social network/Transportation forecasting models
• Agriculture •Stochastic processes
•Supply chain management

DMOT | | Goutam Sutar DMOT | | Goutam Sutar

7 8

Real life examples

• National Broadcast Network Company

• Gov’t owned broadband provider in
Real life examples Australia
• Developed tool to optimize design of
• Kroger
network servicing 8 million locations
• Analytics team created models to determine reorder
point & order up to quantities for items in 1,950 in-store • Benefits:
pharmacies • Reduction in network design time
• Benefits: • Savings of ~$1.7 billion
• Reduced prescription stockouts by 1.6 million
• Lowered inventory by over $120 million
• Increased revenue by $80 million

9 10

Prescriptive Analytics
Real life examples

• Alliance for Paired Donations (APD)

• Transplant patients needing a
kidney often have potential
donors who are incompatible
• APD finds exchanges with other
patient-donor pairs to optimize
paired matches
• Benefits:
• 220 lives saved since 2006
• Value of savings: priceless

11 12

2
 Application and Model Building Application and Model Building

Complexity Level
Definition
“A Optimization problem includes a set of simultaneous linear
equations which represent the condition of the problem and a linear
function which expresses the objective function of the problem”
Therefore, it is also known as Linear Programmimg Problem

DMOT | | Goutam Sutar DMOT | | Goutam Sutar

13 14

Application and Model Building Application and Model Building

Model building approach

Introduction to Linear Programming  Observation - Identification of a problem that exists

(or may occur soon) in a system or organization.
• A Linear Programming model seeks to maximize or minimize a  Problem Definition - The problem must be clearly
and consistently defined, showing its boundaries and
linear function, subject to a set of linear constraints. interactions with the objectives of the organization.
• The linear model consists of the following  Model Construction - Development of the functional
mathematical relationships that describe the decision
components: variables, objective function and constraints of the
– A set of decision variables. problem.
– An objective function.  Model Solution - Models solved using management
science techniques.
– A set of constraints.
 Model Implementation - Actual use of the model or
its solution.

Source: Taylor, 2016 Pearson Education, Inc.

DMOT | | Goutam Sutar DMOT | | Goutam Sutar

15 16

Application and Model Building Application and Model Building

A farmer’s Dilemma
A Profit Maximization Example
A firm produces two products A and B. each unit of product A requires 2 kg
of raw material and 4 labor hours for processing. Similarly, each unit of
product B requires 3 kg of raw material and 3 hours of labor. Let’s assume
that both the products use same type of raw material and equally skilled
human resource. Every week the firm has an availability of 60 kg of raw
material and 96 labor hours. One unit of product A yields Rs 40 and one unit
of product B gives Rs 35 as profit. What will be the best policy?

Answer: A=18, B=8

DMOT | | Goutam Sutar DMOT | | Goutam Sutar

17 18

3
 Application and Model Building Application and Model Building

A farmer’s Dilemma
A farmer has 125 acre farm. He produces radish, muttar and potato. He gets Rs.5 per
Assumptions
kg for radish, Rs.4 per kg for muttar and Rs. 5 per kg for potato. The average per acre
• Assumptions of the linear programming model
yield is 1500 kg of radish, 1800 kg of muttar and 1200 kg of potato. To produce each
100 kg of radish and muttar, a sum of Rs. 12.50 is the manure. Similarly, to produce – The parameter values are known with certainty.
80 kg of potato same sum of Rs 12.50 is the manure. Labor required per acre is 6 – The objective function and constraints exhibit constant returns to scale.
man-days each for radish and potato and 5 man-days for mutter. A total of 500 man- – There are no interactions between the decision variables (the additivity
days of labor is available at the rate of Rs. 40 per man day. assumption).
What will be the best strategy? – The Continuity assumption: Variables can take on any value within a given
feasible range.

Answer: m=100 acres

DMOT | | Goutam Sutar DMOT | | Goutam Sutar

19 20

Application and Model Building Application and Model Building

Formulation of LPP (Most Important)

Frequently Used Terms:
Guidelines:
• Decision Variables
•Understand the problem thoroughly. • Objective function
•Define the decision variables.
• Constraints
•Describe the objective.
• Conditions or Situations
•Describe each constraint.
•Write the objective function in terms of the decision variables. • Optimal Solution
•Write the constraints in terms of the decision variables. • Degeneracy and other issues.

“Correct problem formulation = Problem is almost solved”

DMOT | | Goutam Sutar DMOT | | Goutam Sutar

21 22

Application and Model Building Application and Model Building

Question 1
A manufacturer of a line of patent medicine is preparing a product plan on
medicines A and B. There are enough ingredients available to make 20,000
Problem Formulation bottles A and 40,000 bottles of B but there are only 45,000 bottles into which
either of the medicines can be put. Furthermore, it takes 3 hours to prepare
Correct formulation of real world problem is equivalent to most part enough materials to fill 1000 bottles of A and 1 hour to prepare enough materials
of problem SOLVED. to fill 1000 bottles of B. 66 hours are only available for this operation. The profit is
Rs 8/- per bottle for A and Rs 7/- per bottle of B. Formulate the problem as an
LPP. How should the manufacturer schedule production in order to maximize the
profit.

DMOT | | Goutam Sutar DMOT | | Goutam Sutar

23 24

4
 Application and Model Building Application and Model Building

A farmer’s Dilemma
Question 2: A farmer has 125 acre farm. He produces radish, muttar and potato. He gets Rs.5
A firm can produce three kinds of cloth say A, B, and C. Three kinds of wool are per kg for radish, Rs.4 per kg for muttar and Rs. 5 per kg for potato. The average
required for it i.e., red, green and blue wools. One-unit length of type A cloth needs 2 per acre yield is 1500 kg of radish, 1800 kg of muttar and 1200 kg of potato.To
yards of red wool and 3 yards of blue wool; one-unit length of type B needs 3 yards produce each 100 kg of radish and muttar, a sum of Rs. 12.50 is the manure.
of red, 2 yards of green and 2 yards of blue wools. Type C cloth requires 5 yards of Similarly, to produce 80 kg of potato same sum of Rs 12.50 is the manure. Labour
green and 4 yards of blue wools. The firm has only a stock of 8 yards of red, 10 required per acre is 6 man days each for radish and potato and 5 man days for
yards of green and 15 yards of blue wools. It is assumed that the income obtained mutter. A total of 500 man days of labour is available at the rate of Rs. 40 per man
from one-unit length of type A cloth is $ 3; of type B $5 and of type C is $4. day.
Determine how the firm should use the available material to maximize the profit. What will be the best strategy? Consider yield as the revenue generated by selling
1 kg of the crops.
Answer: m=100 acres

DMOT | | Goutam Sutar DMOT | | Goutam Sutar

25 26

Application and Model Building Application and Model Building

Solution
Let r, m, p be the acreages of land used for radish, mutter and potato respectively.
ion Methodology
Max (5*1500 -187.5 – 6*40) r +
(4*1800 – 225 - 5*40) m + Solution Methodology
(5*1200 – 187.5 – 6*40)p
= Max (7072.5 r +6775.0 m+5572.5 p
s.t r + m+ p <= 125 (Land constraint)
6r + 5m + 6 p <= 500 (Man days constraint) • Graphical Method
r, m, p >= 0 (non-negativity) • Simplex Algorithm

Answer: m=100 acres

DMOT | | Goutam Sutar DMOT | | Goutam Sutar

27 28

Application and Model Building Application and Model Building

The Galaxy Industries Production Problem –

A firm which was famous for assembling computers and computer equipments is about to start production A Prototype Example
of two new web server models. Each type of model requires assembly time, inspection time and
storage space. The design and manufacturing dept. provided the following information. • Galaxy manufactures two toy gun models:
Per unit parameters Type 1 Type 2
– Space Ray.
Profit $ 60 $50 – Zapper.
Assembly time 4 hours 10 hours
Inspection time 2 hours 1 hours
• Resources are limited to
Storage space 3 cubic feet 3 cubic feet – 1000 pounds of special plastic.
The resources available for assembly time is 100 hours, inspection time is 22 hours and the storage – 40 hours of production time per week.
space is 39 cubic space. It is assumed that all the demands can be sold out because of the brand
value. Then, what is the manager’s decision?

DMOT | | Goutam Sutar DMOT | | Goutam Sutar

29 30

5
 Application and Model Building Application and Model Building
The Galaxy Industries Production Problem –
The Galaxy Industries Production Problem – A Prototype Example A Prototype Example

• Marketing requirement
– Total production cannot exceed 700 dozens. • The current production plan calls for:
– Number of dozens of Space Rays cannot exceed – Producing as much as possible of the more profitable product, Space Ray ($8
profit per dozen).
number of dozens of Zappers by more than 350
– Use resources left over to produce Zappers ($5 profit
units. per dozen), while remaining within the marketing guidelines.
• Technological input
• The current production plan consists of:
– Space Rays requires 2 pounds of plastic and
Space Rays = 450 dozen 8(450) + 5(100)
3 minutes of labor per dozen.
Zapper = 100 dozen
– Zappers requires 1 pound of plastic and Profit = $4100 per week
4 minutes of labor per dozen.

DMOT | | Goutam Sutar DMOT | | Goutam Sutar

31 32

Application and Model Building Application and Model Building

Management is seeking a A linear programming model

production plan that will can provide an insightful and an
increase the company’s profit. intelligent solution to this problem.

DMOT | | Goutam Sutar DMOT | | Goutam Sutar

33 34

Application and Model Building Application and Model Building

The Galaxy Linear Programming Model The Galaxy Linear Programming Model
• Decisions variables: Max 8X1 + 5X2 (Weekly profit)
– X1 = Weekly production level of Space Rays (in dozens) subject to
– X2 = Weekly production level of Zappers (in dozens). 2X1 + 1X2  1000 (Plastic)
3X1 + 4X2  2400 (Production Time)
• Objective Function:
X1 + X2  700 (Total production)
– Weekly profit, to be maximized X1 - X2  350 (Mix or difference)
Xj> = 0, j = 1,2 (Nonnegativity)
DMOT | | Goutam Sutar DMOT | | Goutam Sutar

35 36

6
 Application and Model Building Application and Model Building

The Graphical Analysis of Linear

Programming
Using a graphical presentation
The set of all points that satisfy all the we can represent all the constraints,
constraints of the model is called
a
the objective function, and the three

FEASIBLE REGION
types of feasible points.

DMOT | | Goutam Sutar DMOT | | Goutam Sutar

37 38

Application and Model Building Application and Model Building

Graphical Analysis – the Feasible Region Graphical Analysis – the Feasible Region

X2
X2
1000 The Plastic constraint
2X1+X2  1000
700 Total production constraint:
X1+X2  700 (redundant)
The non-negativity constraints 500

Infeasible
Production Feasible
X1 Time
3X1+4X2  2400 X1
500 700

DMOT | | Goutam Sutar DMOT | | Goutam Sutar

39 40

Application and Model Building Application and Model Building

Graphical Analysis – the Feasible Region

X2
1000 The Plastic constraint
2X1+X2 1000
700 Total production constraint:

500
X1+X2 700 (redundant) • There are three
types of
Solving Graphically for an
Infeasible feasible points
Optimal Solution
Production mix
constraint:
Production Feasible X1-X2  350
Time
3X1+4X22400
X1
500 700
Interior points. Boundary points. Extreme points.

DMOT | | Goutam Sutar DMOT | | Goutam Sutar

41 42

7
 Application and Model Building Application and Model Building
The search for an optimal solution

X2 Start at some arbitrary profit, say profit = $2,000... Summary of the optimal solution
1000 Then increase the profit, if possible...
...and continue until it becomes infeasible Space Rays = 320 dozen
Zappers = 360 dozen
700 Profit =$4360 Profit = $4360
500 – This solution utilizes all the plastic and all the production hours.

– Total production is only 680 (not 700).

– Space Rays production exceeds Zappers production by only 40

X1 dozens.
500

DMOT | | Goutam Sutar DMOT | | Goutam Sutar

43 44

Application and Model Building Application and Model Building

Extreme points and optimal solutions Multiple optimal solutions

– If a linear programming problem has an optimal • For multiple optimal solutions to exist, the objective
solution, an extreme point is optimal. function must be parallel to one of the constraints

•Any weighted average of

optimal solutions is also an
optimal solution.

DMOT | | Goutam Sutar DMOT | | Goutam Sutar

45 46