You are on page 1of 17

1.

Introduction to Operations Research

Dr. Foo Chuan Hui| M231 21th Oct 2023

Discover theDiscover
world at the
Leiden
world
University
at Leiden University
Introduction to O.R.

• O.R. is primarily emerged during World War II.

• O.R. was known as Operational Research during


WW2.

• The father of O.R. is often attributed to George Dantzig


who had developed linear programming, a fundamental
technique in operations research

Discover the world at Leiden University 1


Introduction to O.R.

• It is originated in Britain where a team of scientists made maximum use of limited military resources
via quantitative techniques.

• It leads to the optimal solution of the problem which is close to the real solution.

• It is the most effective means in resolving complex problems efficiently (LP, Simulation, Queuing
theory…)

• OR is an analytical method of decision making that is useful in the management of organization such
as businesses and networks of machines.
Examples of machine networks: computer networks, distributed systems, Internet of Things (IoT)

Discover the world at Leiden University 2


Introduction to O.R.
Scope of OR:
OR is known as a multidisciplinary field that applies mathematical and analytical methods to solve
complex problems.
• Business and management.
• Transportation and logistics.
• Healthcare.
• Finance.
• Manufacturing and production.
• Energy and resource management.
• Education etc.

Discover the world at Leiden University 3


Introduction to O.R.
O.R. methodology 1. Formulate
the problem

1. Define the problem clearly and precisely and 2. Construct


establish the objectives through the OR 6. Implement
mathematical
process. the solution
model

2. Express the problem into mathematical


model to show the relationship between the O.R.
parameters: process
- Objective function
- Constraints
- Decision variables 5. Establish
3. Solve the
controls over
model
the solution
3. Three fundamental methods to the solution:
- Analytical method
4. Validate
- Numerical method the model

- Simulation

Discover the world at Leiden University 4


Introduction to O.R.
O.R. methodology 1. Formulate
the problem

4. Comparing the present data with past data, if 6. Implement 2. Construct


the results show the similarity, the given mathematical
the solution
model
model can be accepted.
O.R.
5. Control needs to be setup for detecting the process
changes in the given parameters and
providing feedback whether the changed 5. Establish
3. Solve the
controls over
parameters are accepted or not. the solution
model

6. Monitor the implementation process and 4. Validate


the model
address any issues that arise

Discover the world at Leiden University 5


Mathematical Modeling
Pastry products:
A bakery produces two types of pastries: croissants and muffins. Each pastry is made with a
different recipe, requires different ingredients, and yields a different profit.
Each croissant yields a profit of $2, and each muffin yields a profit of $1. To make one croissant, the
bakery requires 2 cups of flour and 1/4 cup of butter. To make one muffin, the bakery requires 1 cup
of flour and 1/8 cup of butter. There are 20 cups of flour and 4 cups of butter available. The bakery
can make up to 30 pastries in total (croissants and muffins combined).
The bakery wants to determine the best combination of croissants and muffins to maximize profit
while considering constraints.

Discover the world at Leiden University 6


Mathematical Modeling
Pastry products:
A bakery produces two types of pastries: croissants and muffins. Each pastry is made with a
different recipe, requires different ingredients, and yields a different profit.
Each croissant yields a profit of $2, and each muffin yields a profit of $1. To make one croissant, the
bakery requires 2 cups of flour and 1/4 cup of butter. To make one muffin, the bakery requires 1 cup
of flour and 1/8 cup of butter. There are 20 cups of flour and 4 cups of butter available. The bakery
can make up to 30 pastries in total (croissants and muffins combined).
The bakery wants to determine the best combination of croissants and muffins to maximize profit
while considering constraints.

Decision variables:
• x = # of croissants
• y = # of muffins

Discover the world at Leiden University 7


Mathematical Modeling
Pastry products:
A bakery produces two types of pastries: croissants and muffins. Each pastry is made with a
different recipe, requires different ingredients, and yields a different profit.
Each croissant yields a profit of $2, and each muffin yields a profit of $1. To make one croissant, the
bakery requires 2 cups of flour and 1/4 cup of butter. To make one muffin, the bakery requires 1 cup
of flour and 1/8 cup of butter. There are 20 cups of flour and 4 cups of butter available. The bakery
can make up to 30 pastries in total (croissants and muffins combined).
The bakery wants to determine the best combination of croissants and muffins to maximize profit
while considering constraints.

Decision variables: Objective: maximize profit, z


• x = # of croissants z = 2x + y
• y = # of muffins

Discover the world at Leiden University 8


Mathematical Modeling
Pastry products:
A bakery produces two types of pastries: croissants and muffins. Each pastry is made with a
different recipe, requires different ingredients, and yields a different profit.
Each croissant yields a profit of $2, and each muffin yields a profit of $1. To make one croissant, the
bakery requires 2 cups of flour and 1/4 cup of butter. To make one muffin, the bakery requires 1 cup
of flour and 1/8 cup of butter. There are 20 cups of flour and 4 cups of butter available. The bakery
can make up to 30 pastries in total (croissants and muffins combined).
The bakery wants to determine the best combination of croissants and muffins to maximize profit
while considering constraints.

Constraints: Variable domains


• Flour Constraint: 2x + y ≤ 20 Total Pastries Constraint: x + y ≤ 30
• Butter Constraint: ¼ x + 1/8 y ≤ 4 Non-negativity Constraints: x ≥ 0, y ≥ 0

Discover the world at Leiden University 9


Mathematical Modeling
Pastry products:
A bakery produces two types of pastries: croissants and muffins. Each pastry is made with a
different recipe, requires different ingredients, and yields a different profit.
Each croissant yields a profit of $2, and each muffin yields a profit of $1. To make one croissant, the
bakery requires 2 cups of flour and 1/4 cup of butter. To make one muffin, the bakery requires 1 cup
of flour and 1/8 cup of butter. There are 20 cups of flour and 4 cups of butter available. The bakery
can make up to 30 pastries in total (croissants and muffins combined).
The bakery wants to determine the best combination of croissants and muffins to maximize profit
while considering constraints.

We call this a program. It is a linear program, because the objective is a linear function of the
decision variables, and the constraints are linear inequalities (in the decision variables).

Discover the world at Leiden University 10


Mathematical Modeling
Furniture:
A small furniture manufacturer produces two types of furniture: chairs and tables. The profit for
each product is calculated based on the original cost and labor fees. Each chair is sold for $150 and
costs $50 in raw materials, $40 in labor fees, and $10 in overhead expenses. Each table is sold for
$250 and costs $80 in raw materials, $60 in labor fees, and $20 in overhead expenses. The
company has a budget constraint, meaning it can't spend more than $5,000 on raw materials and
labor fees. There's also a space constraint, and the company can produce at most 50 pieces of
furniture.
The company wants to determine the best production plan to maximize profit while considering
resource constraints.

Discover the world at Leiden University 11


Linear Programming

An optimization problem has a standard form:

• We maximize / minimize a linear function of the decision variables. This function is known as
objective function.

• The values of the decision variables must satisfy a set of constraints, each consisting of a linear
equation or linear inequality

• a sign restriction, i.e., usually nonnegative ( xi ≥ 0 ) but perhaps nonpositive ( xi ≤ 0 ), may be


associated with each decision variable.

Discover the world at Leiden University 12


Linear Programming
Example:

Max 2𝑥1 + 𝑥2

Constraint to
3𝑥1 − 𝑥2 ≥ 6
2𝑥1 + 3𝑥2 ≤ 12
𝑥1 , 𝑥2 ≥ 0

Discover the world at Leiden University 13


Linear Programming
Solution:

Max 2𝑥1 + 𝑥2

Constraint to
3𝑥1 − 𝑥2 ≥ 6
2𝑥1 + 3𝑥2 ≤ 12
𝑥1 , 𝑥2 ≥ 0

Each point in the shaded feasible region satisfies all four inequality constraints is called
feasible solution.

Discover the world at Leiden University 14


Linear Programming
Solution:

Max 2𝑥1 + 𝑥2

Constraint to
3𝑥1 − 𝑥2 ≥ 6
2𝑥1 + 3𝑥2 ≤ 12
𝑥1 , 𝑥2 ≥ 0

The optimal solution is the feasible solution for which the objective function is largest.

Discover the world at Leiden University 15


Linear Programming
Solution:

By graphing the linear eq.:

2𝑥1 + 𝑥2 = 0,
2𝑥1 + 𝑥2 = 4,
2𝑥1 + 𝑥2 = 12 𝑒𝑡𝑐.

The optimal solution :

Discover the world at Leiden University 16

You might also like