Linear programming

A linear programming problem (also known as a linear optimization problem) involves a linear
objective function, which is the function that must be maximized or minimized. This objective
function is subject to some constraints, which are inequalities or equations that restrict the values of
the variables. To illustrate these concepts, suppose a manufacturer produces 17- and 24-inch
computer monitors. Past sales experience shows that at least twice as many 17-inch monitors are
sold as 24-inch monitors. Suppose further that the manufacturing plant is capable of producing 12
monitors per day. Let 𝑥 represent the number of 17-inch monitors produced per day, and let 𝑦
represent the number of 24-inch monitors produced per day. Then

These two inequalities place constraints, or restrictions, on the manufacturer.

Further, suppose a profit of $50 is earned on each 17-inch monitor sold and $75 is earned on each
24-inch monitor sold. Then the manufacturer’s daily profit P, in dollars, is given by the equation

The equation 𝑃 = 50𝑥 + 75𝑦 defines the objective function. The goal of this linear programming problem is to
determine how many of each monitor should be produced to maximize the manufacturer’s profit and satisfy the
constraints.

Because the manufacturer cannot produce fewer than zero units of either monitor, there are two
other implied constraints: 𝑥 ≥ 0, 𝑦 ≥ 0. Our linear programming problem now looks like

To solve this problem, we graph the solution set of the constraints. The solution set of the
constraints is called the set of feasible solutions. Ordered pairs in this set are used to evaluate the
objective function to determine which ordered pair maximizes the profit. For example, (5, 2), (8, 3),
(8, 4) and (10, 1) are three ordered pairs in the set (see graph below).

For these ordered pairs, the profits would be

𝑃 = 50(5) + 75(2) = 𝟒𝟎𝟎 (when 𝑥 = 5 and 𝑦 = 2)
𝑃 = 50(8) + 75(3) = 𝟔𝟐𝟓 (when 𝑥 = 8 and 𝑦 = 3)
𝑃 = 50(8) + 75(4) = 𝟕𝟎𝟎 (when 𝑥 = 8 and 𝑦 = 4)
𝑃 = 50(10) + 75(1) = 𝟓𝟕𝟓 (when 𝑥 = 10 and 𝑦 = 1)
It would be impossible to check every ordered pair in the set of feasible solutions to find which one
maximizes profit. The following theorem from linear programming simplifies the task.
In our current example there are only two feasible vertices (corner points): (8, 4) and (12, 0). The
values of the objective function 𝑃 = 50𝑥 + 75𝑦 at these vertices are
𝑃 = 50(8) + 75(4) = 𝟕𝟎𝟎 at (𝑥, 𝑦) = (8, 4)
𝑃 = 50(12) + 75(0) = 𝟔𝟎𝟎 at (𝑥, 𝑦) = (12, 0)
The maximum possible value of the objective function occurs at the point whose coordinates are
(8, 4). The manufacturer will therefore maximize profit by producing 8 17-inch monitors and 4 24-
inch monitors each day. The profit will be $700 per day.

In general, the goal of any linear programming problem is to maximize or minimize the objective
function, subject to the constraints. Minimization problems occur, for example, when a
manufacturer wants to minimize the cost of operations.

Suppose that a cost minimization problem results in the following objective function and constraints.

Below is the graph of the solution set of the constraints.

The task is to find the ordered pair that satisfies all the constraints and gives the smallest value of 𝐶.
The vertices (or corner points) in the feasible region together with the corresponding values of the
objective function 𝐶 = 3𝑥 + 4𝑦 are:

The minimum value of the objective function is 𝐶 = 3, which occurs at the point whose coordinates
are (1, 0).
It is important to realize that the maximum or minimum value of an objective function depends on
the objective function and on the set of feasible solutions. For example, using the set of feasible
solutions in the graph above but changing the objective function to 𝐶 = 2𝑥 + 5𝑦 changes the
maximum value of C to 25, at the ordered pair (0, 5). Verify this result by making a list of vertices
(or corner points) with the corresponding values of the objective function 𝐶, similar to the one
shown above.
Exercise: Minimize the objective function 𝐶 = 4𝑥 + 7𝑦 with the constraints

Now, let us consider some examples ot applied linear programming problems.

Example 1: A manufacturer of animal food makes two grain mixtures, grain 1 (𝐺1) and grain 2
(𝐺2 ). Each mixture contains vitamins, protein, and carbohydrates, in the following proportions:

• Each kilogram of 𝐺1 contains 300 grams of vitamins, 400 grams of protein, and 100 grams of
carbohydrates
• Each kilogram of 𝐺2 contains 100 grams of vitamins, 300 grams of protein, and 200 grams of
carbohydrates

Minimum nutritional guidelines require that a feed mixture made from these grains contain at least
900 grams of vitamins, 2200 grams of protein, and 800 grams of carbohydrate. 𝐺1 costs $2.00 per
kilogram to produce, and 𝐺2 costs $1.25 per kilogram to produce. Find the number of kilograms of
each grain mixture that should be produced to minimize cost.

Solution: Let

The objective function is the cost function

𝐶 = 2𝑥 + 1.25𝑦
Now,
• for the vitamins, because 𝑥 kilograms of 𝐺1 contains 300𝑥 grams of vitamins and 𝑦 kilograms
of 𝐺2 contains 100𝑦grams of vitamins, the total amount of vitamins contained in 𝑥 kilograms
of 𝐺1 and 𝑦 kilograms of 𝐺2 is 300𝑥 + 100𝑦. At least 900 grams of vitamins are necessary, so
300𝑥 + 100𝑦 ≥ 900
• for the protein, following similar reasoning, we have
400𝑥 + 300𝑦 ≥ 2200
• for the carbohydrates, we have
100𝑥 + 200𝑦 ≥ 800
• and lastly, the natural constraints,
𝑥 ≥ 0, 𝑦≥0

The cost minimization problem results in the following objective function

𝐶 = 2𝑥 + 1.25𝑦
Subject to the constraints

The graph of the solution set of the constraints is shown below.

Substituting the coordinates of the vertices into the objective function we get

The minimum value of the objective function is $9.50. It occurs when the company produces a feed
mixture that contains 1 kilogram of 𝐺1 and 6 kilograms of 𝐺2 .

Example 2: A company manufactures two types of cleansers. One is an all-purpose cleanser (AP)
and the other is an industrial strength cleanser (IS). Each cleanser is a mixture of three chemicals, as
follows:
• Each kilolitre of AP requires 12 litres of surfactants, 9 litres of enzymes, and 30 litres of
solvents
• Each kilolitre of IS requires 24 litres of surfactants, 5 litres of enzymes, and 30 litres of
solvents

The profit per kilolitre from the AP cleanser is $100, and the profit per kilolitre from the IS cleanser is
$85. The inventory of the company shows 480 litres of surfactants, 180 litres of enzymes, and 720
litres of solvents available. Assuming the company can sell all the cleanser it produces; find the
number of kilolitres of each cleanser the company should produce to maximize profit. What is the
maximum profit?

Solution: Let

The objective function is the profit function

𝑃 = 100𝑥 + 85𝑦
because 𝑥 kilolitres of AP requires 12𝑥 litres of surfactants and 𝑦 kilolitres of IS requires 24𝑦 litres of
surfactants, the total amount of surfactants needed in kilolitres of AP and 𝑦 kilolitres of IS is 12𝑥 +
24𝑦. There are 480 litres of surfactants in inventory, so
12𝑥 + 24𝑦 ≤ 480
Following similar reasoning, we have the constraints, put together as,

The graph of the solution set of the constraints is as shown below.

Substituting the coordinates of the vertices into the objective function we have

The maximum value of the objective function is $2265 when the company produces 15 kilolitres of
the all-purpose cleanser and 9 kilolitres of the industrial strength cleanser.

Exercises
1. In each of the following cases find the minimum value of the given objective function for the
given set of feasible solutions. Also state the coordinates of the point where the objective
function takes on its minimum value.
2. In each of the following cases find the minimum value of the given objective function for the
given set of feasible solutions. Also state the coordinates of the point where the objective
function takes on its minimum value.

3. A dietician formulates a special breakfast cereal by mixing Oat Flakes (OF) and Crunchy O’s
(CO). The cereals each provide protein and carbohydrates in the following amounts:
• 1 cup of OF provides 6 grams of protein, and 30 grams of carbohydrates
• 1 cup of CO provides 3 grams of protein, and 40 grams of carbohydrates
 The dietician wants to produce a mixture that contains at least 210 grams of protein and at
least 1200 grams of carbohydrates. The cost is 38 cents for 1 cup of Oat Flakes and 32 cents
for 1 cup of Crunchy O’s. How many cups of each cereal will satisfy the constraints and
minimize the cost? What is the minimum cost?

4. A tent manufacturer makes a two-person tent and a family tent. Each type of tent requires
time in the cutting room and time in the assembly room, as indicated below.
• A two-person tent requires 2 hours in the cutting room and 2 hours in the assembly
room
• A family tent requires 2 hours in the cutting room4 hours in the assembly room
The total number of hours available per week in the cutting room is 50. There are 80 hours
available per week in the assembly room. The manager requires that the number of two-
person tents manufactured be no more than four times the number of family tents
manufactured. The profit for the two-person tent is $34, and the profit for the family tent is
$49. Assuming that all the tents produced can be sold, how many of each should be
manufactured per week to maximize the profit? What is the maximum profit?

5. A farmer is planning to raise wheat and barley. Each acre of wheat yields a profit of $50, and
each acre of barley yields a profit of $70. To sow the crop, two machines, a tractor and a
tiller, are rented. The tractor is available for 200 hours, and the tiller is available for 100
hours. Sowing an acre of barley requires 3 hours of tractor time and 2 hours of tilling. Sowing
an acre of wheat requires 4 hours of tractor time and 1 hour of tilling. How many acres of
each crop should be planted to maximize the farmer’s profit?

6. An ice cream supplier has two machines that produce vanilla and chocolate ice cream. The
production rates of each machine are as follows:
• Machine 1 produces 4 gallons of vanilla per hour and 5 gallons of chocolate per hour
• Machine 2 produces 3 gallons of vanilla per hour and 10 gallons of chocolate per
hour
To meet one of its contractual obligations, the company must produce at least 60 gallons of
vanilla ice cream and 100 gallons of chocolate ice cream per day. It costs $28 per hour to run
machine 1 and $25 per hour to run machine 2. How many hours should each machine be
operated to fulfill the contract at the least expense?