📘 Lecture 45 — Planning Production Levels: Linear Programming
📖 Overview: This lecture introduces Linear Programming (LP) as a mathematical method for optimizing resource allocation under constraints. It covers the core components of LP models, their assumptions, and demonstrates how to formulate and solve production planning problems using graphical analysis.
🗂️ Topics Covered
The lecture reviews the components of a linear programming model including decision variables, objective functions, and constraints. It presents the methodology for formulating LP problems, discusses the importance and assumptions of LP, and works through a detailed production problem example involving two doll models. The dentist practice allocation problem is introduced as another application. Graphical analysis is explained step-by-step for finding feasible regions and optimal solutions, including the concepts of extreme points and multiple optimal solutions.
📝 Lecture Summary
INTRODUCTION TO LINEAR PROGRAMMING
A Linear Programming model seeks to maximize or minimize a linear function, subject to a set of linear constraints. The linear model consists of: a set of decision variables (xⱼ), an objective function (Σcⱼxⱼ), and a set of constraints (Σaᵢⱼxⱼ ≤ bᵢ).
THE FORMAT FOR AN LP MODEL
Maximize or minimize Σcⱼxⱼ = c₁x₁ + c₂x₂ + ... + cₙxₙ Subject to: aᵢⱼxⱼ ≤ bᵢ, for i = 1,...,m Non-negativity conditions: all xⱼ ≥ 0, j = 1,...,n
Here n is the number of decision variables and m is the number of constraints (there is no relation between n and m).
THE METHODOLOGY OF LINEAR PROGRAMMING
- Define decision variables
- Hand-write objective
- Formulate math model of objective function
- Hand-write each constraint
- Formulate math model for each constraint
- Add non-negativity conditions
THE IMPORTANCE OF LINEAR PROGRAMMING
Many real world problems lend themselves to linear programming modeling and can be approximated by linear models. There are well-known successful applications in: Operations, Marketing, Finance (investment), Advertising, and Agriculture. The output from linear programming packages provides useful “what if” analysis.
ASSUMPTIONS OF THE LINEAR PROGRAMMING MODEL
- The parameter values are known with certainty
- The objective function and constraints exhibit constant returns to scale
- There are no interactions between the decision variables (the additivity assumption)
- The Continuity assumption: Variables can take on any value within a given feasible range
💡 Why this matters: These assumptions form the foundation of LP. Violating them means the LP model may not accurately represent the real problem.
A PRODUCTION PROBLEM – A PROTOTYPE EXAMPLE
A company manufactures two toy doll models: Doll A and Doll B. Resources are limited to: 1000 kg of special plastic, 40 hours of production time per week. Marketing requirement: Total production cannot exceed 700 dozens, and the number of dozens of Model A cannot exceed the number of dozens of Model B by more than 350.
Decisions variables: X₁ = Weekly production level of Model A (in dozens) X₂ = Weekly production level of Model B (in dozens)
Objective Function: Maximize 800X₁ + 500X₂ (Weekly profit)
Subject to: 2X₁ + 1X₂ ≤ 1000 (Plastic) 3X₁ + 4X₂ ≤ 2400 (Production Time) X₁ + X₂ ≤ 700 (Total production) X₁ - X₂ ≤ 350 (Mix) Xⱼ ≥ 0, j = 1,2 (Nonnegativity)
ANOTHER EXAMPLE
A dentist is faced with deciding how best to split his practice between general dentistry and pedodontics (children’s dental care). The dentist employs three assistants and uses two operatories.
Each pedodontic service requires: .75 hours of operatory time, 1.5 hours of an assistant’s time, and .25 hours of the dentist’s time.
A general dentistry service requires: .75 hours of an operatory, 1 hour of an assistant’s time, and .5 hours of the dentist’s time.
Net profit: Rs. 1000 for each pedodontic service and Rs. 750 for each general dental service.
Time each day: eight hours of dentist’s, 16 hours of operatory time, and 24 hours of assistants’ time.
THE GRAPHICAL ANALYSIS OF LINEAR PROGRAMMING
Using a graphical presentation, we can represent all the constraints, the objective function, and the three types of feasible points.
GRAPHICAL ANALYSIS – THE FEASIBLE REGION: The feasible region is defined using non-negativity constraints. Each constraint is represented by a straight line, and the feasible region is the intersection area where all constraints are satisfied.
THE SEARCH FOR AN OPTIMAL SOLUTION
Constraints intersect to form a point that represents the optimal solution. The procedure is to start with a point representing a certain profit level (e.g., Rs. 200,000), then move the objective function line upwards (for maximization) until the last point on the feasible region is reached.
SUMMARY OF THE OPTIMAL SOLUTION
Model A = 320 dozen Model B = 360 dozen Profit = Rs. 436,000
This solution utilizes all the plastic and all the production hours. Total production is only 680 (not 700).
EXTREME POINTS AND OPTIMAL SOLUTIONS
If a linear programming problem has an optimal solution, an extreme point is optimal. The optimal solution will always occur at a corner point of the feasible region.
MULTIPLE OPTIMAL SOLUTIONS
There may be more than one optimal solution when the objective function is parallel to one of the constraints. If a weighted average of different optimal solutions is obtained, it is also an optimal solution.
⭐ Key Takeaways
Linear Programming is a powerful optimization tool for allocating limited resources under constraints. The decision variables (X₁, X₂) represent quantities to decide, the objective function maximizes or minimizes a linear expression (e.g., profit), and constraints represent resource limits (e.g., ≤) or other restrictions. The graphical method identifies the feasible region defined by all constraints, and the optimal solution lies at an extreme point (corner) of this region. The prototype production example demonstrates transforming a business problem into a mathematical model and solving it to find optimal production levels (320 dozen of Model A, 360 dozen of Model B) yielding maximum profit of Rs. 436,000.
🧠 Quick Revision Questions
- What are the three core components of any Linear Programming model?
- In the prototype production problem, what do the coefficients 800 and 500 represent in the objective function?
- What does the constraint X₁ - X₂ ≤ 350 represent in the doll production problem?
- Why must the optimal solution always occur at an extreme point of the feasible region?
- Under what condition can there be multiple optimal solutions in a Linear Programming problem?