Introduction

Optimal planning techniques such as linear programming have a plethora of applications ranging from
economics to engineering and military affairs. The idea of linear programming revolves around the idea
of maximizing or minimizing something based on a model with linear functions. This project seeks to
democratize linear programming and its concepts, applications, and techniques further.

Background

The concept of linear programming can be traced back to the early 20th as a result of an algorithm called
the simplex which was invented by a mathematician called George Dantzig in the 1940s. Such a
revolution changed the way everything concerned with problem solving in operations research worked.
Many years later, it is still proven that LP is a very versatile method as it focuses on optimization of
production sequences, logistics and so much more as a way of aiding in decision making.

Objectives

Below are the targets of this project:

Basic Concepts of Linear Programming: Outline the basic concepts regarding LP.

Application of LP: Put to test the applicability of LP in real life situations.

Solution Techniques of LP: Assess solution techniques of LP problems.

Constraints of LP: Investigate the constraints and restrictions associated with LP.

Background Knowledge

Before understanding the general structure of linear programming, there are some key points that
should be emphasized:

Objective Function: One can view this as a mathematical representation, that expresses aims such as
maximizing profits or minimizing expenditures.
Constraints: These can be considered delta challenges that are put on the possible solutions which may
be in relation to other factors like the given resources or funds.

Feasible Region: All possible answers that do not violate the constraints would locate at its vertices and
should be the most likely point for optimal solution.

Basic Feasible Solution: This is the solution which meets all the restrictions and is most of the time
located at the point which is formed by the intersections of the restrictions within the feasible region.

The concepts outlined above are also of particular importance in the evaluation of findings, in relation to
LP, that have been drawn.

Methodology

A linear programming problem consists of several problem areas and is solved in some steps such as:

1.Problem formulation:

The objective function is to be advanced in a unambiguous manner.

All the limitative variables must be indicated and clearly outlined.

There has to be quite clear non-negativity constraints on the decision variables.

2.Graphical representation (to be used in case of two variable problems):

Too many unique weaknesses should go to the nomination on graphs as well as the unique area that
constrains the decision best.

The best zone is the right decision area that the above-mentioned can be practiced upon.

3. Finding the Solution to the LP Problem

Traditionally, in small problems: graphical methods.

However, for larger problems, use the simplex algorithm.

The only reasonable way to do so is to recognize an optimal solution as a value of an objective function
which determines the greatest of all corner points.

4. Sensitivity Analysis:

Interpret slightly the sensitivity of an optimal solution to changes presented by the coefficients of the
objective function and constraints thereby offering an extra dimension of insight into the sensitivity of
the model paradigm.

5. Application:

Make these conclusions in actual practice that recommendations given can be applicable and
appropriate to a specified organization.

Proof: A Case Study

A Case of Maximizing production in a Manufacturing Company

For a manufacturing company, which is producing products A and B, both of them are employed to profit
in different ways. Resources also require balance regarding labor and materials supply.

1. Objective Function:

Maximize

In that case, z = 40A + 30B

3. Constraints:

1. Labour constraint: ÷-)2A + B ≤ 100

2. Material constraint: A + 2B ≤ 80

Graphing all of these constraints will give the feasible region.

The optimal solution may lead the firm to decide that it should
produce 30 units of good A and 20 units of good B to maximize
profit.

Solution
We solve it using the simplex method or graphically to get
maximum profit as follows:
nz = 40(30) + 30(20) = 1200 + 600 = 1800
]
### Code Example
Here is how you could translate this into Python, using the PuLP
library:
```python
# Importing the PuLP library
from pulp import LpProblem, LpMaximize, LpVariable, lpSum,
LpStatus, value

# Create a linear programming problem

problem = LpProblem("Maximize_Profit", LpMaximize)

# Decision Variables
A = LpVariable('A', lowBound=0, cat='Continuous') # Quantity
of Product A
B = LpVariable('B', lowBound=0, cat='Continuous') # Quantity
of Product B

# Objective Function: Maximize z = 40A + 30B

problem += 40 * A + 30 * B, "Total_Profit"

# Constraints
problem += 2 * A + B <= 100, "Labor_Constraint"
problem += A + 2 * B <= 80, "Material_Constraint"

# Solve the problem

problem.solve()

# Output the results

print(f"Status: {LpStatus[problem.status]}
")
print(f"Optimal quantity of Product A: {value(A)}
")
print(f"Optimal quantity of Product B: {value(B)}

")

print(f"Maximum Profit: {value(problem.objective)}

")

Maximum Profit: 1800.0

LIMITATIONS OF LINEAR PROGRAMMING

Though linear programming has a number of advantages as already discussed, it is still

accompanied with a few drawbacks in the following respects:
1 . \textbf{Linearity Assumption: } It is inherent in LP that all data has to be assumed to be linear,
which is in reality not an unreal assumption in many of the problems 2

2. \\ 2. \textbf{Concern for Certainty: } LP has a need for certainty with respect to parameters and data.
Such a level of certainty can never be determined by anything from the real world.

3. \\ 3. \textbf{Integer Constraints: } The majority of practical problems which have a requirement for the
solution to be an integer are inmost cases not solvable with realism directly applying LP.

4. \textbf{Computational Complexity: } The more comprehensive the problem, the more calculating
intensive and hence the more probability of it being impractical to resolve via real world techniques.

5. \textbf{Static Environment Modeling: } It is apparent that LP has been developed within an

environment that is static as thus, a change of that nature over time must not occur.

Conclusion

Linear programming has become increasingly recognized as the single most effective technique available
to optimizing resources towards strategic implementation in numerous fields. Be it owing to the fact that
through the nature of application of a scientific approach toward the solving of complex problems,
efficiency and profitability are without a doubt on the rise. Thereby, having a grasp of principles and
methodologies geared towards understanding it would enable the decision maker to harness LP to great
effect and extricate himself from an operational conundrum.

Conclusion / Direct Future Directions

Further work on hybrid models that combine the advantages of linear programming and integer
programming with the advantages of stochastic optimization methods would be promising. These will
remove the drawbacks of LP and extend its use to more complicated scenarios and that will help in the
improvement of decision making tools. It may combine LP with machine learning and artificial
intelligence in achieving adaptive optimization where the models are automatically reconfigured
according to the environment.