Finding the Best Value: Applied Extrema
Optimization problems ask: what is the maximum area, minimum cost, or optimal design given real-world constraints? The strategy is straightforward: set up two equations—an objective function (the quantity being optimized) and a constraint (a fixed relationship between variables)—then reduce to a single-variable problem solvable by calculus.
The Core Method
Every optimization has two essential equations:
| Component | Purpose | Example |
|---|---|---|
| Objective function | Quantity to maximize/minimize | Area, volume, cost, distance |
| Constraint | Fixed relationship between variables | Fixed perimeter, budget, or volume |
Solve the constraint for one variable and substitute into to eliminate variables. This produces , a function of a single variable ready for critical-point analysis on a domain determined by physical meaning (e.g., ).
Worked Example: Maximizing Rectangular Area
A farmer has 100 meters of fencing. Maximize the enclosed area. Let sides be and .
Constraint:
Objective: for
Optimize: . Since , this is a maximum. The optimal field is a square with area .
Key Pitfalls to Avoid
Forgetting the domain: Physical constraints (lengths positive, volumes positive) must restrict . In the farming example, , not all real numbers. Endpoints represent degenerate shapes and yield zero or nonsensical values.
Skipping reasonableness checks: Always verify that your critical point makes physical sense. Negative dimensions or impossible values signal setup errors.
The Optimization Checklist
- Identify objective quantity in terms of two (or more) variables
- Write the constraint equation linking those variables
- Solve constraint for one variable; substitute into
- Determine domain from physical restrictions
- Find critical points using ; apply derivative test or closed-interval method
- Verify the answer is physically reasonable
Mastering steps 1–2—translating a word problem into correct equations—is the genuine skill. Once is established on its proper domain, calculus delivers the answer.