Introduction to Linear Programming (LP)

Linear Programming (LP) is a mathematical technique used for optimization—finding the

best possible outcome given a set of constraints. It is widely used in fields such as business,
economics, engineering, logistics, and operations research.

1. Definition of Linear Programming

Linear Programming is a method used to maximize or minimize a linear objective function
subject to a set of linear constraints.

Objective Function – A mathematical expression that represents the goal of the problem (e.g.,
maximizing profit, minimizing cost).
Constraints – A set of linear inequalities or equations that limit the possible solutions.
Decision Variables – The unknowns we are solving for, which impact the objective function.
Feasible Region – The set of all possible solutions that satisfy the constraints.

2. Importance of Linear Programming

 Helps in decision-making and resource optimization.
 Widely used in business, economics, and engineering.
 Solves complex problems like supply chain management, production scheduling, and
financial planning.

3. Components of a Linear Programming Problem

To understand Linear Programming, let's break it down into its components with an example.

Example: A company produces two products (A and B) and wants to maximize profit. The
company has limited resources in terms of labor and raw materials.

Decision Variables:

 Let x1 be the number of units of Product A produced.

 Let x2 be the number of units of Product B produced.

Objective Function (Maximize Profit):

Maximize Z=5x1+3x2
(Where $5$ and $3$ are the profit per unit for products A and B, respectively.)
🔹 Constraints (Resource Limitations):
Labor Constraint: 2x1+x2≤100
Material Constraint: x1+2x2≤80
Non-negativity Constraint: x1,x2≥0
(Since we cannot produce negative quantities.)
4. Methods to Solve Linear Programming Problems

There are two main methods to solve LP problems:

 Graphical Method – Used when there are only two decision variables. The solution is
found by plotting constraints on a graph and identifying the feasible region.
 Simplex Method – Used when there are more than two decision variables. It involves
iterative calculations using a tableau.

5. Applications of Linear Programming

Real-World Uses:

 Business & Economics: Profit maximization, cost minimization, pricing strategies.

 Manufacturing: Production scheduling, resource allocation, material blending.
 Transportation & Logistics: Route optimization, delivery scheduling, warehouse
management.
 Finance & Investment: Portfolio optimization, risk minimization.
 Agriculture: Crop planning, land use optimization, fertilizer allocation.

Conclusion

Linear Programming is a powerful tool for optimizing resources and making efficient decisions
under constraints. It provides practical solutions to real-world problems in various industries.
Basics of the Simplex Method
The Simplex Method is an iterative algorithm used to solve Linear Programming (LP)
problems that involve more than two variables. It systematically finds the optimal solution by
moving along the edges of the feasible region.

1. What is the Simplex Method?

 It is a mathematical approach used to solve LP problems when the Graphical Method is not
feasible (i.e., when there are more than two decision variables).
 It helps in maximizing or minimizing an objective function subject to a set of linear
constraints.
 The method uses a tableau (matrix form) to perform systematic calculations.

2. When is the Simplex Method Used?

 When there are more than two decision variables.
 When a problem has many constraints, making the graphical approach impractical.
 When the feasible region is large, and an efficient way to find the optimal solution is
needed.

3. Components of a Linear Programming Problem in the

Simplex Method
A standard LP problem for the Simplex Method consists of:

 Decision Variables – Variables that determine the outcome (e.g., x1,x2,x3).

 Objective Function – The function to maximize or minimize:
Maximize Z=c1x1+c2x2+...+cnxn
 Constraints – Linear inequalities or equations limiting the values of the decision variables:
a11x1+a12x2+...+a1nxn≤b1
a21x1+a22x2+...+a2nxn≤b2
 Non-Negativity Condition – Ensures all variables remain non-negative:
x1,x2,...,xn≥0

4. Steps of the Simplex Method

The Simplex Method follows a structured approach:

Step 1: Convert the Problem to Standard Form

 Ensure all constraints are in the form of equations by introducing slack variables.

Example: Convert 3x1+5x2≤10 into 3x1+5x2+s1=10, where s1s_1s1 is a slack variable.

Step 2: Set Up the Initial Simplex Tableau

 Construct a tableau (matrix) to represent the system of equations.

Step 3: Identify the Pivot Column (Entering Variable)

 Choose the column with the most negative value in the objective row (if maximizing).
 This column indicates which variable should enter the solution.

Step 4: Identify the Pivot Row (Leaving Variable)

 Compute the minimum ratio test to determine which variable exits the solution.
 The smallest positive ratio determines the pivot row.

Step 5: Perform Row Operations (Pivoting)

 Normalize the pivot row to make the pivot element 1.

 Adjust the rest of the tableau using row operations to make all other values in the pivot column
zero.

Step 6: Repeat Until Optimal Solution is Found

 Check if all values in the objective row (except for the right-hand side) are non-negative.
 If any values are still negative, repeat steps 3-5.
 When all values in the objective row are non-negative, the optimal solution is reached.

5. Example of the Simplex Method

Problem Statement:
Maximize

Z=3x1+2x2
Subject to:
x1+x2≤4
2x1+x2≤5
x1,x2≥0

Step 1: Convert to Standard Form

Introduce slack variables s1 and s2:

x1+x2+s1=4

2x1+x2+s2=5

Objective Function:

Z−3x1−2x2=0

Step 2: Construct the Initial Tableau

Basis x1 x2 s1 s2 RHS
s1 1 1 1 0 4
s2 2 1 0 1 5
Z -3 -2 0 0 0

(Continue with pivoting steps until all negative values in the Z-row are removed.)
The Simplex Algorithm Process
The Simplex Algorithm is an iterative method used to solve Linear Programming (LP) problems. It
efficiently finds the optimal solution by systematically improving the objective function value through
tableau operations.

1. Key Components of the Simplex Algorithm

Before we begin the process, let's recall the main components of an LP problem:

 Objective Function: The function to maximize or minimize, such as:

Maximize Z=c1x1+c2x2+...+cnxn\text{Maximize } Z = c_1x_1 + c_2x_2 + ... +
c_nx_nMaximize Z=c1x1+c2x2+...+cnxn
 Constraints: A set of linear inequalities or equations limiting the variables.
 Decision Variables: Variables that influence the solution.
 Feasible Region: The set of all possible solutions satisfying the constraints.

2. The Step-by-Step Simplex Algorithm

The Simplex Method follows these structured steps:

Step 1: Convert the Problem into Standard Form

To apply the Simplex Method, the LP problem must be in standard form, meaning:

 The objective function is in maximization form.

 The constraints are in equality form (≤ constraints require slack variables).
 All variables are non-negative (xi≥0).

Example: Convert the following LP problem into standard form.

Given Problem:
Maximize:
Z=3x1+2x2
Subject to:
x1+x2≤4
2x1+x2≤5
x1,x2≥0

Introduce slack variables s1s_1s1 and s2s_2s2 to convert inequalities into equations:
x1+x2+s1=4
2x1+x2+s2=5
Z−3x1−2x2=0

Step 2: Construct the Initial Simplex Tableau

The tableau represents the system in matrix form:

Basis x1 x2 s1 s2 RHS
s1 1 1 1 0 4
 Basis x1 x2 s1 s2 RHS
s2 2 1 0 1 5
Z -3 -2 0 0 0

 Basic Variables (Basis): Initially, slack variables s1and s2 form the basis.
 Right-Hand Side (RHS): Represents the available resources.
 Objective Row (Z-row): Contains coefficients from the objective function.

Step 3: Identify the Pivot Column (Entering Variable)

The pivot column is the one with the most negative value in the Z-row (objective function
row).

 In our case, the most negative coefficient is -3 (under x1), so x1 enters the basis.

Step 4: Identify the Pivot Row (Leaving Variable)

To find which variable exits the basis:

 Compute the Minimum Ratio Test:

Ratio=RHS/Pivot Column

Row RHS x1x_1x1 Ratio

s1 4 1 4/1=4
s2 5 2 5/2=2.5

The smallest positive ratio is 2.5, so s2 leaves the basis.

Step 5: Perform Row Operations (Pivoting)

1. Make the Pivot Element = 1

 The pivot element is 2 (in the x1, s2 row).

 Divide the entire second row by 2.

2. Make Other Entries in Pivot Column = 0

 Use row operations to make other elements in the x1x_1x1 column zero.

New Tableau after pivoting:

Basis X1 X2 S1 S2 RHS
S1 0 0.5 1 -0.5 1
X1 1 0.5 0 0.5 2.5
Z 0 -0.5 0 1.5 7.5

Step 6: Repeat Until Optimal Solution is Found

 The new Z-row still has a negative value (-0.5 in x2x_2x2), so we repeat steps 3-5.
 We select x2x_2x2 as the entering variable and perform another round of pivoting.
  Once all values in the Z-row (except RHS) are non-negative, we have reached the optimal
solution.

Step 7: Interpret the Final Tableau

Once all values in the Z-row are non-negative, the optimal solution is found.

From the final tableau:

 Optimal Values: x1=2.5x_1 = 2.5x1=2.5, x2=3x_2 = 3x2=3, and Maximum ZZZ = 7.5.
 Non-basic variables (slack variables) indicate unused resources.

3. Special Cases in the Simplex Algorithm

🔹 Unbounded Solution: If no limit exists for increasing the objective function.
🔹 Infeasible Solution: If no feasible region satisfies all constraints.
🔹 Degeneracy: Occurs when a basic variable takes a zero value, leading to cycling.
🔹 Multiple Optimal Solutions: Happens when the objective function is parallel to a
constraint.
 Quantitative Method
Activities/Exercises

General Instructions for Solving Simplex Method Exercises

1. Identify the Objective Function

o Determine if the problem is Maximization or Minimization.
o Rewrite it in standard form: Z=c1x1+c2x2+⋯+cnxn

2. Convert Constraints into Equations

o If a constraint is ≤, introduce a slack variable si to convert it into an equation.
o If a constraint is ≥, introduce a surplus variable si and an artificial variable ai
(use Big M Method or Two-Phase Method).
o Rewrite the system in augmented form.

3. Set Up the Initial Simplex Tableau

o Create a tableau with:
 Decision variables x1,x2,...x_1, x_2, ...
 Slack or artificial variables
 Objective function row (Z-row)

4. Choose the Entering Variable (Pivot Column)

o Identify the most negative value in the Z-row.
o The corresponding column is the pivot column.

5. Choose the Leaving Variable (Pivot Row)

o Apply the minimum ratio test:

Right-Hand Side (RHS)/Pivot Column Value

o The smallest non-negative ratio determines the pivot row.

6. Perform Row Operations

o Make the pivot element 1 by dividing the pivot row.
o Make all other values in the pivot column zero by using row operations.

7. Repeat Steps 4-6 Until the Z-row Has No Negative Values

o If all values in the Z-row are non-negative, the solution is optimal.

8. Interpret the Final Tableau

o The values in the Basic Variable (BV) column give the optimal values for the
decision variables.
o The objective function value (Z) is found in the RHS of the Z-row.

Specific Instructions for Each Exercise

Exercise 1: Basic Maximization Problem

Objective: Maximize Z=5x1+3x2

 Introduce slack variables s1,s2

 Solve using Simplex Tableau
 Find optimal x1,x2

Exercise 2: Another Maximization Example

Objective: Maximize Z=6x1+4x2

 Add slack variables

 Use pivot column and row selection
 Perform row operations
 Find the optimal solution

Exercise 3: Minimization Problem

Objective: Minimize Z=2x1+3x2

 Convert minimization into maximization by multiplying Z by −1

 Add surplus and artificial variables
 Use Big M Method or Two-Phase Method
 Solve using Simplex Table

Exercise 4: Degeneracy in Simplex Method

Objective: Maximize Z=3x1+5x2

 Follow the standard Simplex Method steps

 Check if any basic variable becomes zero during iterations

Exercise 5: Multiple Optimal Solutions

Objective: Maximize Z=4x1+2x2

 Solve using Simplex Method

 If there is a zero in the Z-row for a non-basic variable, check for alternative optimal
solutions

Final Notes:

 Use row operations carefully.

 Always double-check pivot column and pivot row selection.
 If an artificial variable is still in the final solution, the problem has no feasible
solution.
Graphical Method Exercises in Linear Programming

Below are a few exercises that require solving Linear Programming (LP) problems using the
Graphical Method. Each exercise includes constraints, an objective function, and step-by-step
instructions.

Exercise 1: Basic Graphical Maximization Problem

Problem Statement:
Maximize:

Z=3x1+5x2

Subject to:

2x1+x2≤8
x1+2x2≤6
x1,x2≥0

Instructions:

1. Plot Constraints:
o Convert inequalities into equations.
o Find intercepts by setting x1=0 and solving for x2, and vice versa.
o Draw lines for constraints on a graph.
o Identify the feasible region where all constraints are satisfied.

2. Find Corner Points:

o Determine intersection points of constraint lines.
o Identify extreme points of the feasible region.

3. Compute Z for Each Corner Point:

o Evaluate the objective function at each feasible vertex.
o The maximum Z-value gives the optimal solution.

Exercise 2: Another Maximization Problem

Maximize:

Z=4x1+6x2

Subject to:

x1+x2≤5
x1+3x2≤9
x1,x2≥0

Follow the Same Steps as Exercise 1

1. Convert inequalities into lines.

2. Plot them on a graph and shade the feasible region.
3. Identify feasible corner points.
 4. Compute ZZ at each vertex and choose the maximum.

Exercise 3: Minimization Problem

Minimize:

Z=2x1+3x2

Subject to:

3x1+x2≥6
x1+2x2≥4
x1,x2≥0

Instructions:

 Convert inequalities into ≥ equations.

 Plot constraints and shade the feasible region.
 Find feasible corner points.
 Compute Z at each vertex and choose the minimum Z-value.

Exercise 4: Unbounded Feasible Region

Maximize:

Z=5x1+2x2

Subject to:

x1−x2≤4
x1+x2≥3
x1,x2≥0

Instructions:

 Follow previous steps to plot constraints.

 Observe if the feasible region extends indefinitely.
 If Z keeps increasing, the solution is unbounded.

Scoring Rubric for Each Exercise

Task Points

Convert inequalities into equations 3

Correctly plot lines and feasible region 4

Identify feasible corner points 3

Compute Z-values correctly 3

Identify correct optimal solution 2

Total 15 points