4.

6 Optimization
In this section, our goal is to be able to do the following:

1. Relate the concept of Absolute Extrema to Optimization

2. Use Absolute Extrema in the procedure of solving optimization problems

3. Deduce the function to be optimized given an optimization problem

4. Solve optimization problems of different types:

(a) Maximizing area, perimeter, or similar problems

(b) Maximizing revenue or profit

Day 1: Introduction and Problems involving Area or Perimeter

From the name itself, optimization deals with finding the highest or lowest values of certain
attributes in a given problem. One example of a problem involving optimization is when we’re
given a budget for a rectangular fence, and we want to maximize the area while staying within
the budget. In this particular problem, we want to optimize the ”area” attribute given the
constraint of the budget for the fence. Another possible optimization problem is when we’re
given the price-demand equation of a particular product as well as its cost function. For this
scenario, a possible problem to be solved is how many items to produce in order to maximize
revenue or profit.

When we want to optimize a certain attribute such as area or profit, we usually take the absolute
extrema of the function which models that particular attribute. Recall that absolute extrema
gives us the largest or smallest values of a given function over an interval, if they exist. Hence,
it is logical to use these absolute extrema when solving for optimization problems. Later on
when we do optimization problems, this will be clear.

There are many different types of optimization problems ranging from minimizing the cost of
piping across a river and the riverbank to analyzing the cost-effectiveness of flying different
airline routes. However, for the purposes of this course as well as in the interest of time, we
will only be tackling two types of optimization problems:

1. Area and Perimeter Problems 2. Revenue or Profit Problems

In many optimization problems, we aren’t given any functions immediately from the statement
of the problem. We are then required to deduce the function to be optimized as well as any
constraints to be factored into the optimization problem. Thus, this is an essential skill when
solving these types of problems.
To solve optimization problems, we present a general procedure to follow. Even though we
have a step-by-step process to follow, it’s also important to understand why we do these steps
instead of just memorizing the procedure.

Procedure: Strategy for Solving Optimization Problems

1. Introduce variables, look for relationships among the variables, and

construct a mathematical model of the form

Max (or min) f (x) on the interval I

2. Find the critical numbers of f (x).

3. Find the relevant absolute extrema of f (x) on the interval I and the
x’s where this occurs.

4. Use the solution to the mathematical model to answer all the questions
asked in the problem.

Note that in the procedure above, f (x) is to be only a function of one variable. In the case
where you get a function to be optimized and it is a function of two or more variables, you
must be able to transform this function into a function of only one variable by exploring the
various constraints to the problem.

First, we try our hand at area and perimeter problems, more specifically, maximizing area.

Example 1: Maximizing Area (P. 304)

A homeowner has $320 to spend on building a fence around a rectangular

garden. Three sides of the fence will be constructed with wire fencing at a
cost of $2 per linear foot. The fourth side will be constructed with wood
fencing at a cost of $6 per linear foot. Find the dimensions and the area of
the largest garden that can be enclosed with $320 worth of fencing.

Try this problem out first before moving on. You should get the following. The dimensions of
the garden with the maximum area of 800 ft.2 are 20 ft. x 40 ft., with one 20-foot side of wood
fencing.
The area of the rectangular garden is given by product of the lengths of the height and width
of the rectangle, which gives us A = xy, where x is the width of the rectangle and y is the
height. Without loss of generality, we assume that the side with the wood fencing is on the
width. This means that the total cost of producing the fence is given by:

Cost = 2y + 2x + 2y + 6x = 8x + 4y = 320

This gives us the optimization problem:

Max A = xy, subject to 8x + 4y = 320

What we want to do next is to make the function to be maximized into a function of only one
variable, say x. From 8x + 4y = 320, we get y = 80 − 2x, which we can substitute into the
original optimization function:

Max A(x) = x(−2x + 80) = −2x2 + 80x

To maximize the area A(x), we must first get the domain of the function. We know that the
sides of a rectangle must be positive, so we get x ≥ 0 and y ≥ 0. From y ≥ 0, we can get
x ≤ 40:

y = 80 − 2x ≥ 0 ⇒ 2x ≤ 80 ⇒ x ≤ 40

This gives us the final domain of x ∈ [0, 40] or 0 ≤ x ≤ 40. Next, we want to find the critical
numbers, and we can get this by getting the zeroes of the derivative of the area function A(x):

A0 (x) = −4x + 80 = 0 ⇒ x = 20

Thus, our singular critical number is x = 20. Finally, we want to evaluate A(x) on the critical
number x = 20 and the end points of the domain: x = 0 and x = 40.
Substituting the critical numbers and end points:

x A(x)
0 0
20 800
40 0

Here, we see that the highest value of the area is when x = 20, which gives a total area of 800
ft.2 . Since y = 80 − 2x, substituting x = 20 gives us y = 80 − 2(20) = 40.

Thus, the dimensions of the rectangle which maximizes area is 20 ft. x 40 ft. with one of the 20
ft. sides being fenced by wood. This gives us the maximized area of 800 ft.2 with $320 worth
of fencing.

Note: In the illustration above, we coincidentally set x to be the longer side of the rectangle
and y to be the shorter side. When we got the dimensions of the rectangle which optimized the
area, we got x = 20 and y = 40. Even though this is contradictory to the original illustration
we made, it does not matter. The illustration only serves as a guide to set up the necessary
functions and does not affect the final answer.

If you want to see this problem solved in a video, click on the link below:

https://youtu.be/bn5hRtfCQMM

Next, we do an example on minimizing perimeter.

Example 2: Minimizing Perimeter (P. 306)

Refer to Example 1. The homeowner judges that an area of 800 square

feet for the garden is too small and decides to increase the area to 1,250
square feet. What is the minimum cost of building a fence that will enclose
a garden with an area of 1,250 square feet? What are the dimensions of this
garden? Assume that the cost of fencing remains unchanged.

Like with the first example, try this problem out first before moving on. You should get
the following. The minimum cost for enclosing a 1,250-square foot garden is $400, and the
dimensions are 25 ft x 50 ft, with one 25-foot side of wood fencing.
This extension of the first example puts a spin on the procedure from the first problem. In
example 1, the area was the function to be optimized and cost of the fencing was the constraint.
Now, things are the other way around; the cost of the fencing is the function to be optimized
and the area is the constraint:

Cost = 8x + 4y, and Area = xy = 1250

This gives us the optimization problem:

Min C = 8x + 4y, subject to xy = 1250

Similar to Example 1, we want to make the function to be minimized into a function of only
1250
one variable, say x. From xy = 1250, we get y = , which we can substitute into the
x
original optimization function:
 
1250 5000
Min C(x) = 8x + 4 = 8x +
x x

Like Example 1, we must first get the domain of the function C(x). This time, since our
constraint is xy = 1250, neither x nor y can be equal to zero, so we get x > 0 and y > 0.

1250
y>0 ⇒ >0
x

1250
> 0 will be true only for positive values of x. Thus, this gives us the final domain of
x
x ∈ (0, +∞). Next, we want to find the critical numbers. We get the derivative C 0 (x) and find
the critical numbers:

5000
C 0 (x) = 8 − ⇒ x = 25
x2

5000
From 8 − , we’re actually supposed to get 3 critical numbers: x = −25, 0, 25, but since our
x2
domain is (0, +∞), we take only x = 25.

Since our domain isn’t a closed interval, we can’t use the same method as in Example 1 to find
the absolute extrema. We will have to use the second derivative test on our singular critical
number to determine if it is an absolute minimum:

10, 000 10, 000

C 00 (x) = ⇒ C 00 (25) = = 0.64 > 0
x3 253

Since C 00 (25) > 0, x = 25 is an absolute minimum. This gives us the minimum cost of
5000 1250 1250
C(25) = 8(25) + = $400. Since y = , substituting x = 25 gives us y = = 50.
25 x 25
Thus, the dimensions of the rectangle which minimizes cost is 25 ft. x 50 ft. with one of the 25
ft. sides being fenced by wood. This gives us the minimized cost of $400 with an area of 1250
ft.2 .