Linear Programming Chapter: Systems of Linear Inequalities

332600_09_1.
qxp 4/22/08 4:50 PM Page 531
9 Linear Programming
9.1 Systems of Linear CHAPTER OBJECTIVES

Inequalities
■ Sketch the graph of a linear inequality.
9.2 Linear Programming
Involving Two Variables ■ Sketch the graph of a system of linear inequalities.
9.3 The Simplex Method: ■ Sketch the solution of a system of linear inequalities.
Maximization ■ Derive a set of inequalities that represents a system of linear inequalities.
9.4 The Simplex Method: ■ Find a maximum or minimum of an objective function subject to a system of constraints.
Minimization ■ Sketch a graph of the solution region for a linear programming problem.
9.5 The Simplex Method:
■ Write the simplex tableau for a linear programming problem.
Mixed Constraints
■ Use the simplex method to solve a linear programming problem that maximizes an objective
function.
■ Determine the dual of a linear programming problem that minimizes an objective function.
■ Use the simplex method to solve a linear programming problem that minimizes an objective
function.
■ Add the appropriate slack variables to, or subtract the appropriate surplus variables from,
a system and form the initial simplex tableau.
■ Use specified entering and departing variables to solve a mixed-constraint problem.
9.1 Systems of Linear Inequalities

The following two statements are inequalities in two variables.
3x ⫺ 2y < 6 and x ⫹ y ⱖ 6
An ordered pair 共a, b兲 is a solution of an inequality in x and y if the inequality is true when
a and b are substituted for x and y, respectively. For instance, 共1, 1兲 is a solution of the
inequality 3x ⫺ 2y < 6 because 3共1兲 ⫺ 2共1兲 ⫽ 1 < 6. The graph of an inequality is the
collection of all solutions of the inequality. To sketch the graph of an inequality such as
3x ⫺ 2y < 6,
begin by sketching the graph of the corresponding equation
3x ⫺ 2y ⫽ 6.
531
332600_09_1.qxp 4/22/08 4:50 PM Page 532
532 Chapter 9 Linear Programming
The graph of the equation will normally separate the plane into two or more regions.
In each such region, one of the following two statements must be true. (1) All points in the
region are solutions of the inequality. (2) No points in the region are solutions of
the inequality. So, you can determine whether the points in an entire region satisfy the
inequality by simply testing one point in the region.
Sketching the Graph 1. Replace the inequality sign with an equal sign, and sketch the graph of the resulting
of an Inequality equation. 共We use a dashed line for < or > and a solid line for ⱕ or ⱖ .兲
2. Test one point in each of the regions formed by the graph in step 1. If the point
in Two Variables satisfies the inequality, then shade the entire region to denote that every point in the
region satisfies the inequality.
In this section, you will work with linear inequalities of the form
ax ⫹ by < c ax ⫹ by ⱕ c
ax ⫹ by > c ax ⫹ by ⱖ c.
The graph of each of these linear inequalities is a half-plane lying on one side of the line
ax ⫹ by ⫽ c. The simplest linear inequalities are those corresponding to horizontal or
vertical lines, as shown in Example 1.
EXAMPLE 1 Sketching the Graph of a Linear Inequality

Sketch the graphs of (a) x > ⫺2 and (b) y ⱕ 3.
SOLUTION (a) The graph of the corresponding equation x ⫽ ⫺2 is a vertical line. The points that
satisfy the inequality x > ⫺2 are those lying to the right of this line, as shown in
Figure 9.1.
(b) The graph of the corresponding equation y ⫽ 3 is a horizontal line. The points
that satisfy the inequality y ⱕ 3 are those lying below (or on) this line, as shown in
Figure 9.2.
y y
x = −2
y=3
3
2 2
1 1
x x
−3 −1 1 2 3 −3 −2 −1 1 2 3
−2 −2
−3 −3
Figure 9.1 Figure 9.2

332600_09_1.qxp 4/22/08 4:50 PM Page 533
Section 9.1 Sy stems of Linear Inequalities 533
EXAMPLE 2 Sketching the Graph of a Linear Inequality

Sketch the graph of x ⫺ y < 2.
SOLUTION The graph of the corresponding equation x ⫺ y ⫽ 2 is a line, as shown in Figure 9.3.
Because the origin 共0, 0兲 satisfies the inequality, the graph consists of the half-plane lying
above the line. (Try checking a point below the line. Regardless of which point you choose,
you will see that it does not satisfy the inequality.)
y
2
x−y=2
1
(0, 0)
x
−3 −2 −1 1 2 3
−2
−3
Figure 9.3
For a linear inequality in two variables, you can sometimes simplify the graphing proce-
dure by writing the inequality in slope-intercept form. For instance, by writing x ⫺ y < 2
in the form
y > x ⫺ 2,
you can see that the solution points lie above the line y ⫽ x ⫺ 2, as shown in Figure 9.3.
Similarly, by writing the inequality 3x ⫺ 2y > 5 in the form
3 5
y < x⫺ ,
2 2
you see that the solutions lie below the line y ⫽ 32x ⫺ 52.
Systems of Inequalities
Many practical problems in business, science, and engineering involve systems of linear
inequalities. An example of such a system is shown below.
x⫹y ⱕ 12
3x ⫺ 4y ⱕ 15
x ⱖ 0
y ⱖ 0
332600_09_1.qxp 4/22/08 4:50 PM Page 534
A solution of a system of inequalities in x and y is a point 共x, y兲 that satisfies each

inequality in the system. For instance, 共2, 4兲 is a solution of this system because x ⫽ 2 and
y ⫽ 4 satisfy each of the four inequalities in the system. The graph of a system of
inequalities in two variables is the collection of all points that are solutions of the system.
For instance, the graph of the system above is the region shown in Figure 9.4. Note that the
point 共2, 4兲 lies in the shaded region because it is a solution of the system of inequalities.
To sketch the graph of a system of inequalities in two variables, first sketch the graph of
each individual inequality (on the same coordinate system) and then find the region that is
common to every graph in the system. For systems of linear inequalities, it is helpful to find
the vertices of the solution region, as shown in Example 3.
y
12
x + y ≤ 12
3x − 4y ≤ 15
9
x≥0
y≥0
6
(2, 4)
3
x
6 9 12
(2, 4) is a solution because it
satisfies the system of inequalities.
Figure 9.4
EXAMPLE 3 Solving a System of Inequalities

Sketch the graph (and label the vertices) of the solution set of the system shown below.
x⫺y < 2
x > ⫺2
y ⱕ 3
SOLUTION You have already sketched the graph of each of these inequalities in Examples 1 and 2. The
triangular region common to all three graphs can be found by superimposing the graphs on
the same coordinate plane, as shown in Figure 9.5. To find the vertices of the region, we find
the points of intersection of the boundaries of the region.
Vertex A: 共⫺2, ⫺4兲 Vertex B: 共5, 3兲 Vertex C: 共⫺2, 3兲
Obtained by finding Obtained by finding Obtained by finding
the point of the point of the point of
intersection of intersection of intersection of
x ⫺ y ⫽ ⫺2 x⫺y⫽2 x ⫽ ⫺2
x ⫽ ⫺2. y ⫽ 3. y ⫽ ⫺3.
332600_09_1.qxp 4/22/08 4:50 PM Page 535
y
x−y<2 y x > −2
6 6
4 C = (− 2, 3) B = (5, 3)
2 2
x x
−4 2 4 6 8 −6 −4 2 4 6
−2 −2
−4
A = (− 2, − 4)
−6 −6
y≤3
Figure 9.5
For the triangular region shown in Figure 9.5, each point of intersection of a pair of
boundary lines corresponds to a vertex. With more complicated regions, two border lines
can sometimes intersect at a point that is not a vertex of the region, as shown in Figure 9.6.
In order to keep track of which points of intersection are actually vertices of the region,
make a careful sketch of the region and refer to your sketch as you find each point of
intersection.
When solving a system of inequalities, you should be aware that the system might have
no solution. For instance, the system
x⫹y > 3
x ⫹ y < ⫺1
has no solution points because the quantity 共x ⫹ y兲 cannot be both less than ⫺1 and greater
than 3, as shown in Figure 9.7.
y y
(Not a vertex)
3
x+y>3
2
x 1
x
−1 1 2 3
Border lines can intersect x + y < −1

at a point that is not a vertex. No Solution
332600_09_1.qxp 4/22/08 4:50 PM Page 536
Another possibility is that the solution set of a system of inequalities can be unbounded.
For instance, the solution set of
x⫹ y < 3
x ⫹ 2y > 3
forms an infinite wedge, as shown in Figure 9.8.
y
3 x + 2y > 3
1
x+y<3
x
−1 1 2 3
Unbounded Region
Figure 9.8
Application
The last example in this section shows how a system of linear inequalities can arise in an
applied problem.
EXAMPLE 4 An Application of a System of Inequalities

The liquid portion of a diet is to provide at least 300 calories, 36 units of vitamin A, and 90
units of vitamin C daily. A cup of dietary drink X provides 60 calories, 12 units of vitamin
A, and 10 units of vitamin C. A cup of dietary drink Y provides 60 calories, 6 units of
vitamin A, and 30 units of vitamin C. Set up a system of linear inequalities that describes
the minimum daily requirements for calories and vitamins.
SOLUTION Let
x ⫽ number of cups of dietary drink X and
y ⫽ number of cups of dietary drink Y.
Then, to meet the minimum daily requirements, the inequalities listed below must be
satisfied.
For calories: 60x ⫹ 60y ⱖ 300
For vitamin A: 12x ⫹ 6y ⱖ 36
332600_09_1.qxp 4/22/08 4:50 PM Page 537
For vitamin C: 10x ⫹ 30y ⱖ 90

x ⱖ 0
y ⱖ 0
The last two inequalities are included because x and y cannot be negative. The graph of this
system of linear inequalities is shown in Figure 9.9.
y
10
8
(0, 6)
6
(1, 4)
4
(3, 2)
2
(9, 0)
x
2 4 6 8 10
Figure 9.9
Any point inside the shaded region (or on its boundary) shown in Figure 9.9 meets the
minimum daily requirements for calories and vitamins. For instance, 3 cups of dietary drink
X and 2 cups of dietary drink Y supply 300 calories, 48 units of vitamin A, and 90 units of
vitamin C.
SECTION 9.1 Exercises

In Exercises 1–6, match the linear inequality with its graph. [The (c) y (d) y
graphs are labeled (a)–(f).]
4
1. x > 3 2. y ≤ 2 3. 2x ⫹ 3y ≤ 6 2
3
y
4. 2x ⫺ y ≥ ⫺2 5. x ≥ 6. y > 3x
2 1
1
(a) y (b) y
x x
−2 −1 1 2 −2
2
1
1
x
x −2 −1 1 2
1 2 3
−2
332600_09_1.qxp 4/22/08 4:50 PM Page 538
(e) y (f) y 33. ⫺3x ⫹ 2y < 6 34. x ⫺ 7y > ⫺36

x ⫹ 4y > ⫺2 5x ⫹ 2y > 5
2
2
2x ⫹ y < 3 6x ⫺ 5y > 6
1
1
35. x ⱖ 1 36. x ⫹ y < 10
x x
−2 −1 1 2
1 2 4
x ⫺ 2y ⱕ 3 2x ⫹ y > 10
−1 3x ⫹ 2y ⱖ 9 x⫺y < 2
−2 x⫹ y ⱕ 6
−2
37. 2x ⫹ y > 2 38. x ⫺ 7y > ⫺36

In Exercises 7–22, sketch the graph of the linear inequality. 6x ⫹ 3y < 2 5x ⫹ 2y > 5
7. x ≥ 2 8. x ≤ 4 9. y ≥ ⫺1 6x ⫺ 5y > 6
10. y ≤ 3 11. y < 2⫺x 12. y > 2x ⫺ 4 39. ⫺3x ⫹ 2y < 6 40. x ⫺ 2y < ⫺6
13. 2y ⫺ x ≥ 4 14. 5x ⫹ 3y ≥ ⫺15 x ⫺ 4y > ⫺2 5x ⫺ 3y > ⫺9
15. y ≤ x 16. 3x > y 17. y ≥ 4 ⫺ 2x 2x ⫹ y < 3
18. y ≤ 3⫹x 19. 3y ⫹ 4 ≥ x 20. 6 ⫺ 2y < x In Exercises 41–44, derive a set of inequalities that describes
21. 4x ⫺ 2y ≤ 12 22. y ⫹ 3x > 6 the region.
In Exercises 23–26, determine whether each ordered pair is a 41. y 42. y
solution of the system of linear inequalities. 6
4
23. x ⱖ ⫺4 (a) 共0, 0兲 (b) 共⫺1, ⫺3兲 3 4
y > ⫺3 (c) 共⫺4, 0兲 (d) 共⫺3, 11兲 2
2
y ⱕ ⫺8x ⫺ 3
x
1
24. ⫺2x ⫹ 5y ⱖ 3 (a) 共0, 2兲 (b) 共⫺6, 4兲 1 4 6
x −2
y < 4 (c) 共⫺8, ⫺2兲 (d) 共⫺3, 2兲 1 2 3 4
⫺4x ⫺ 2y < 7
43. y 44. y
25. x ⱖ 1 (a) 共0, 1兲 (b) 共1, 3兲
10
y ⱖ 0 (c) 共2, 2兲 (d) 共2, 1兲 8
y ⱕ 2x ⫹ 1 8 6
6 4
26. x ⱖ 0 (a) 共0, ⫺2兲 (b) 共2, 0兲
4 2
y ⱖ 0 (c) 共3, 1兲 (d) 共0, ⫺1兲
2 x
y ⱕ 4x ⫺ 2 2 4 6 8
x −2
−2 2 6 8
In Exercises 27–40, sketch the graph of the solution of the system
of linear inequalities.
In Exercises 45–48, derive a set of inequalities that describes
27. x ⱖ 0 28. x ⱖ ⫺1 29. x⫹ y ⱕ 1
the region.
y ⱖ 0 y ⱖ ⫺1 ⫺x ⫹ y ⱕ 1
x ⱕ 2 x ⱕ 1 y ⱖ 0 45. Rectangular region with vertices at 共2, 1兲, 共5, 1兲, 共5, 7兲, and
y ⱕ 4 y ⱕ 2 共2, 7兲.
30. 3x ⫹ 2y < 6 31. x ⫹ y ⱕ 5 32. 2x ⫹ y ⱖ 2 46. Parallelogram with vertices at 共0, 0兲, 共4, 0兲, 共1, 4兲, and 共5, 4兲.
x > 0 x ⱖ 2 x ⱕ2 47. Triangular region with vertices at 共0, 0兲, 共5, 0兲, and 共2, 3兲.
y > 0 y ⱖ 0 y ⱕ 1 48. Triangular region with vertices at 共⫺1, 0兲, 共1, 0兲, and 共0, 1兲.
332600_09_1.qxp 4/22/08 4:50 PM Page 539
49. A furniture company can sell all the tables and chairs it 54. Rework Exercise 53 using minimum daily requirements of 280
produces. Each table requires 1 hour in the assembly center and units of calcium, 160 units of iron, and 180 units of vitamin B.
113 hours in the finishing center. Each chair requires 112 hours in 55. A small company that manufactures two models of exercise
the assembly center and 112 hours in the finishing center. The machines has an order for 15 units of the standard model and
company’s assembly center is available 12 hours per day, and 16 units of the deluxe model. The company has trucks of two
its finishing center is available 15 hours per day. If x is the different sizes that can haul the products, as shown in the table.
number of tables produced per day and y is the number of
chairs, find a system of inequalities describing all possible
Truck Standard Deluxe
production levels. Sketch the graph of the system.
50. A store sells two models of a certain brand of laptop comput- Large 6 3
ers. Because of the demand, it is necessary to stock at least
twice as many units of the Pro Series as units of the Deluxe Medium 4 6
Series. The costs to the store of the two models are $80 and
$1200, respectively. The management does not want more than Find and graph a system of inequalities describing the numbers
$20,000 in computer inventory at any one time, and it wants of trucks of each size that are needed to deliver the order.
at least four Pro Series computers and two Deluxe Series
computers in inventory at all times. Devise a system of 56. A person’s maximum heart rate is 220 ⫺ x, where x is the per-
inequalities describing all possible inventory levels, and sketch son’s age in years for 20 ⱕ x ⱕ 70. When a person exercises, it
the graph of the system. is recommended that the person strive for a heart rate that is at
least 50% of the maximum and at most 75% of the maximum.
51. A person plans to invest no more than $20,000 in two different
(Source: American Heart Association)
interest-bearing accounts. Each account is to contain at least
$5000. Moreover, one account should have at least twice (a) Write a system of inequalities that describes the region
the amount that is in the other account. Find a system of corresponding to these heart rate recommendations.
inequalities describing the various amounts that can be (b) Sketch a graph of the region in part (a).
deposited in each account, and sketch the graph of the system. (c) Find two solutions of the system and interpret their mean-
52. Two types of tickets are to be sold for a concert. General ings in the context of the problem.
admission tickets cost $15 and grandstand tickets cost $25. The True or False? In Exercises 57 and 58, determine whether each
promoter of the concert must sell at least 15,000 tickets statement is true or false. If a statement is true, give a reason or cite
including 8000 of the $15 general admission tickets and 4000 an appropriate statement from the text. If a statement is false,
of the $25 grandstand tickets. Moreover, the gross receipts must provide an example that shows the statement is not true in all cases
total at least $275,000 in order for the concert to be held. Find or cite an appropriate statement from the text.
a system of inequalities describing the different numbers of
tickets that can be sold, and sketch the graph of the system. 57. A solution of a system of inequalities in x and y is a point 共x, y兲
that satisfies at least one of the inequalities in the system.
53. A dietitian is asked to design a special diet using two different
foods. Each ounce of food X contains 20 units of calcium, 15 58. Two border lines can sometimes intersect at a point that is not
units of iron, and 10 units of vitamin B. Each ounce of food Y a vertex of the region but is the solution of the system of
contains 10 units of calcium, 10 units of iron, and 20 units of inequalities.
vitamin B. The minimum daily requirements in the diet are 300
units of calcium, 150 units of iron, and 200 units of vitamin B.
Find a system of inequalities describing the different amounts
of food X and food Y that can be used in the diet. Sketch the
graph of the system.
332600_09_2.qxp 4/22/08 4:51 PM Page 540
9.2 Linear Programming Involving Two Variables

y Many applications in business and economics involve a process called optimization, which
is used to find the minimum cost, the maximum profit, or the minimum use of resources.
In this section, one type of optimization problem called linear programming is discussed.
A two-dimensional linear programming problem consists of a linear objective function
and a system of linear inequalities called constraints. The objective function gives the
Feasible
solutions quantity that is to be maximized (or minimized), and the constraints determine the set of
feasible solutions.
For example, consider a linear programming problem in which you are asked to
maximize the value of
x z ⫽ ax ⫹ by Objective function
The objective function has its subject to a set of constraints that determines the region indicated in Figure 9.10. Because
optimal value at one of the vertices every point in the region satisfies each constraint, it is not clear how to go about finding the
of the region determined by the point that yields a maximum value of z. Fortunately, it can be shown that if there is an
constraints. optimal solution, it must occur at one of the vertices of the region. In other words,
Figure 9.10 you can find the maximum value by testing z at each of the vertices, as illustrated in
Example 1.
THEOREM 9.1 If a linear programming problem has an optimal solution, it must occur at a vertex of the
Optimal Solution of a Linear set of feasible solutions. If the problem has more than one optimal solution, then at least
Programming Problem one of them must occur at a vertex of the set of feasible solutions. In either case, the value
of the objective function is unique.
EXAMPLE 1 Solving a Linear Programming Problem

Find the maximum value of
z ⫽ 3x ⫹ 2y Objective function
subject to the constraints listed below.

x ≥ 0
y ≥ 0
Constraints
x ⫹ 2y ≤ 4
x⫺ y ≤ 1
332600_09_2.qxp 4/22/08 4:51 PM Page 541
Section 9.2 Linear Programming Involving Two Variables 541
SOLUTION The constraints form the region shown in Figure 9.11. At the four vertices of this region, the
objective function has the values listed below.
y At 共0, 0兲: z ⫽ 3共0兲 ⫹ 2共0兲 ⫽ 0
4 At 共1, 0兲: z ⫽ 3共1兲 ⫹ 2共0兲 ⫽ 3
At 共2, 1兲: z ⫽ 3共2兲 ⫹ 2共1兲 ⫽ 8 (Maximum value of z)
3 At 共0, 2兲: z ⫽ 3共0兲 ⫹ 2共2兲 ⫽ 4
x + 2y = 4
(0, 2) So, the maximum value of z is 8, and this occurs when x ⫽ 2 and y ⫽ 1.
x=0
1 (2, 1)
R E M A R K : In Example 1, try testing some of the interior points in the region. You will see
x−y=1
x that the corresponding values of z are less than 8.
(0, 0) (1, 0) 2 3
y=0 To see why the maximum value of the objective function in Example 1 must occur at a
vertex, consider writing the objective function in the form
Figure 9.11
3 z
y⫽⫺ x⫹ .
2 2
This equation represents a family of lines, each of slope ⫺3兾2. Of these infinitely many
lines, you want the one that has the largest z-value, while still intersecting the region
determined by the constraints. In other words, of all the lines whose slope is ⫺3兾2, you
want the one that has the largest y-intercept and intersects the specified region, as shown in
Figure 9.12. It should be clear that such a line will pass through one (or more) of the
vertices of the region.
(0, 2) y = − 32 x + 2z
(2, 1)
1
x
(0, 0) (1, 0) 2
Figure 9.12
The graphical method for solving a linear programming problem is outlined as follows.
332600_09_2.qxp 4/22/08 4:51 PM Page 542
Graphical Method of To solve a linear programming problem involving two variables by the graphical method,
Solving a Linear use the following steps.
Programming Problem 1. Sketch the region corresponding to the system of constraints. (The points inside or on
the boundary of the region are called feasible solutions.)
2. Find the vertices of the region.
3. Test the objective function at each of the vertices and select the values of the variables
that optimize the objective function. For a bounded region, both a minimum and
maximum value will exist. (For an unbounded region, if an optimal solution exists, it
will occur at a vertex.)
These guidelines will work regardless of whether the objective function is to be

maximized or minimized. For instance, in Example 1 the same test used to find the
maximum value of z can be used to conclude that the minimum value of z is 0, and this
occurs at the vertex 共0, 0兲.
EXAMPLE 2 Solving a Linear Programming Problem

Find the maximum value of the objective function
where x ⱖ 0 and y ⱖ 0, subject to the constraints

⫺x ⫹ y ⱕ 11
x ⫹ y ⱕ 27 Constraints
2x ⫹ 5y ⱕ 90.
SOLUTION The region bounded by the constraints is shown in Figure 9.13. By testing the objective
function at each vertex, you obtain
At 共0, 0兲: z ⫽ 4共0兲 ⫹ 6共0兲 ⫽ 0
At 共0, 11兲: z ⫽ 4共0兲 ⫹ 6共11兲 ⫽ 66
At 共5, 16兲: z ⫽ 4共5兲 ⫹ 6共16兲 ⫽ 116
At 共27, 0兲: z ⫽ 4共27兲 ⫹ 6共0兲 ⫽ 108.
So, the maximum value of z is 132, and this occurs when x ⫽ 15 and y ⫽ 12.
In the next example, you can see that the same basic procedure can be used to solve a
linear programming problem in which the objective function is to be minimized.
332600_09_2.qxp 4/22/08 4:51 PM Page 543
y y
− x + y = 11 (1, 5)
25 5
2x + 5y = 90
20 4
(5, 16) (0, 4) (6, 3)
15 (15, 12) 3
10 (0, 11) x + y = 27 (0, 2)
2
5 1
(0, 0) (27, 0)
x x
5 10 15 20 25 30 1 2 (3, 0) 4 (5, 0) 6
EXAMPLE 3 Minimizing an Objective Function

Find the minimum value of the objective function

2x ⫹ 3y ⱖ 6
3x ⫺ y ⱕ 15
Constraints
⫺x ⫹ y ⱕ 4
2x ⫹ 5y ⱕ 27.
SOLUTION The region bounded by the constraints is shown in Figure 9.14. By testing the objective
function at each vertex, you obtain
At 共0, 2兲: z ⫽ 5共0兲 ⫹ 7共2兲 ⫽ 14 (Minimum value of z)
At 共0, 4兲: z ⫽ 5共0兲 ⫹ 7共4兲 ⫽ 28
At 共1, 5兲: z ⫽ 5共1兲 ⫹ 7共5兲 ⫽ 40
At 共6, 3兲: z ⫽ 5共6兲 ⫹ 7共3兲 ⫽ 51
At 共5, 0兲: z ⫽ 5共5兲 ⫹ 7共0兲 ⫽ 25
At 共3, 0兲: z ⫽ 5共3兲 ⫹ 7共0兲 ⫽ 15.
So, the minimum value of z is 14, and this occurs when x ⫽ 0 and y ⫽ 2.
R E M A R K : In Example 3, note that the steps used to find the minimum value are
precisely the same ones you would use to find the maximum value. In other words, once
you have evaluated the objective function at the vertices of the feasible region, simply
choose the largest value as the maximum and the smallest value as the minimum.
332600_09_2.qxp 4/22/08 4:51 PM Page 544
y When solving a linear programming problem, it is possible that the maximum (or mini-
(0, 4) (2, 4) mum) value occurs at two different vertices. For instance, at the vertices of the region
4
shown in Figure 9.15, the objective function
3 z = 12 for any
point along z ⫽ 2x ⫹ 2y Objective function
2 this line. has the following values.
1 (5, 1)
(0, 0) At 共0, 0兲: z ⫽ 2共0兲 ⫹ 2共0兲 ⫽ 0
x
1 2 3 4 (5, 0) At 共0, 4兲: z ⫽ 2共0兲 ⫹ 2共4兲 ⫽ 8
Figure 9.15 At 共2, 4兲: z ⫽ 2共2兲 ⫹ 2共4兲 ⫽ 12 (Maximum value of z)
At 共5, 0兲: z ⫽ 2共5兲 ⫹ 2共0兲 ⫽ 10
In this case, you can conclude that the objective function has a maximum value (of 12) not
only at the vertices 共2, 4兲 and 共5, 1兲, but also at any point on the line segment connecting
these two vertices.
Some linear programming problems have no optimal solution. This can occur if the
region determined by the constraints is unbounded. Example 4 illustrates such a problem.
EXAMPLE 4 An Unbounded Region


x ⫹ 2y ⱖ 4
3x ⫹ y ⱖ 7 Constraints
⫺x ⫹ 2y ⱕ 7.
SOLUTION The region determined by the constraints is shown in Figure 9.16. For this unbounded region,
y
there is no maximum value of z. To see this, note that the point 共x, 0兲 lies in the region
5
for all values of x ⱖ 4. By choosing large values of x, you can obtain values of
(1, 4) z ⫽ 4共x兲 ⫹ 2共0兲 ⫽ 4x that are as large as you want. So, there is no maximum value of z.
4
1 (2, 1)
x
1 2 3 (4, 0) 5
Figure 9.16
332600_09_2.qxp 4/22/08 4:51 PM Page 545
Application
EXAMPLE 5 An Application: Minimum Cost

In Example 4 in Section 9.1, you set up a system of linear equations for the following
problem. The liquid portion of a diet is to provide at least 300 calories, 36 units of
vitamin A, and 90 units of vitamin C daily. A cup of dietary drink X provides 60 calories,
12 units of vitamin A, and 10 units of vitamin C. A cup of dietary drink Y provides 60
calories, 6 units of vitamin A, and 30 units of vitamin C. Now, suppose that dietary drink
X costs $.12 per cup and drink Y costs $.15 per cup. How many cups of each drink should
be consumed each day to minimize the cost and still meet the stated daily requirements?
SOLUTION Begin by letting x be the number of cups of dietary drink X and y be the number of cups of
dietary drink Y. Moreover, to meet the minimum daily requirements, the inequalities listed
below must be satisfied.
For calories: 60x ⫹ 60y ⱖ 300
For vitamin A: 12x ⫹ 6y ⱖ 36
For vitamin C: 10x ⫹ 30y ⱖ 90 Constraints
x ⱖ 0
y ⱖ 0
y
The cost C is represented by
10 C ⫽ 0.12x ⫹ 0.15y. Objective function
8 The graph of the region corresponding to the constraints is shown in Figure 9.17. To
(0, 6) determine the minimum cost, test C at each vertex of the region, as follows.
6
(1, 4) At 共0, 6兲: C ⫽ 0.12共0兲 ⫹ 0.15共6兲 ⫽ 0.90
4
At 共1, 4兲: C ⫽ 0.12共1兲 ⫹ 0.15共4兲 ⫽ 0.72
(3, 2)
2 At 共3, 2兲: C ⫽ 0.12共3兲 ⫹ 0.15共2兲 ⫽ 0.66 (Minimum value of C)
(9, 0)
x
At 共9, 0兲: C ⫽ 0.12共9兲 ⫹ 0.15共0兲 ⫽ 1.08
2 4 6 8 10
So, the minimum cost is $.66 per day, and this occurs when three cups of drink X and two
Figure 9.17 cups of drink Y are consumed each day.
332600_09_2.qxp 4/22/08 4:51 PM Page 546

In Exercises 1–16, find the minimum and maximum values of the 7. Objective function: 8. Objective function:
objective function, subject to the indicated constraints. z ⫽ 25x ⫹ 30y z ⫽ 16x ⫹ 18y
1. Objective function: 2. Objective function: Constraints: Constraints:
z ⫽ 3x ⫹ 2y z ⫽ 4x ⫹ 3y (See Exercise 5.) (See Exercise 6.)
Constraints: Constraints: 9. Objective function: 10. Objective function:
x ⱖ 0 x ⱖ 0 z ⫽ 4x ⫹ 5y z ⫽ 4x ⫹ 5y
y ⱖ 0 y ⱖ 0 Constraints: Constraints:
x ⫹ 3y ⱕ 15 2x ⫹ 3y ⱖ 6 x ⱖ 0 x ⱖ 0
4x ⫹ y ⱕ 16 3x ⫺ 2y ⱕ 9 y ⱖ 0 y ⱖ 0
x ⫹ 5y ⱕ 20 4x ⫹ 3y ⱖ 27 2x ⫹ 2y ⱕ 10
y y x⫹ y ⱖ 8 x ⫹ 2y ⱕ 6
(0, 5) 3x ⫹ 5y ⱖ 30
5 5
(3, 4) (0, 4) 11. Objective function: 12. Objective function:
4 4
(5, 3) z ⫽ 2x ⫹ 7y z ⫽ 2x ⫺ y
3 3
(0, 2) Constraints: Constraints:
2 2
(See Exercise 9.) (See Exercise 10.)
1 1 (3, 0)
(0, 0) (4, 0) 13. Objective function: 14. Objective function:
x x
1 2 3 4 5 1 2 3 4 5 z ⫽ 4x ⫹ y z⫽x
z ⫽ 5x ⫹ 0.5y z ⫽ x ⫹ 6y x ⱖ 0 x ⱖ 0
Constraints: Constraints: y ⱖ 0 y ⱖ 0
(See Exercise 1.) (See Exercise 2.) x ⫹ 2y ⱕ 40 2x ⫹ 3y ⱕ 60
x ⫹ y ⱖ 30 2x ⫹ y ⱕ 28
5. Objective function: 6. Objective function:
2x ⫹ 3y ⱖ 72 4x ⫹ y ⱕ 48
z ⫽ 10x ⫹ 7y z ⫽ 50x ⫹ 35y
0 ⱕ x ⱕ 60 x ⱖ 0 z ⫽ x ⫹ 4y z⫽y
0 ⱕ y ⱕ 45 y ⱖ 0 Constraints: Constraints:
5x ⫹ 6y ⱕ 420 8x ⫹ 9y ⱕ 7200 (See Exercise 13.) (See Exercise 14.)
8x ⫹ 9y ⱖ 5400 In Exercises 17–20, maximize the objective function subject to the
y y
constraints 3x1 ⫹ x2 ⱕ 15 and 4x1 ⫹ 3x2 ⱕ 30, where x1, x2 ⱖ 0.
(0, 800)
60
800 17. z ⫽ 2x1 ⫹ x2 18. z ⫽ 5x1 ⫹ x2
(0, 45) (30, 45) 600 19. z ⫽ x1 ⫹ x2 20. z ⫽ 3x1 ⫹ x2
40 400
(60, 20) (900, 0)
200
20 (0, 600)
x
(0, 0) 200 (675, 0)
x
20 40 (60, 0)
332600_09_2.qxp 4/22/08 4:51 PM Page 547
21. An online electronics website plans to sell two models of home 26. A manufacturer produces two models of snowboards. The
computers at costs of $250 and $400. The $250 model yields a Rookie model requires 2.5 hours of assembly, while the Pro
profit of $45 and the $400 model yields a profit of $50. The model requires 3.0 hours of assembly. The Rookie model also
website estimates that the total monthly demand will not requires 2.0 hours of prep and painting while the Pro model
exceed 250 units. Find the number of units of each type of requires 4.0 hours. Both the Rookie and Pro models require
computer that should be stocked in order to maximize profit. only one hour for packaging. The total amounts of time
Assume that the website does not want to invest more than available for assembly, painting, and packaging are 4000 hours,
$70,000 in computer inventory. 5000 hours, and 1500 hours, respectively. The profits per unit
22. A fruit grower has 150 acres of land available to raise two are $50 for the Rookie model and $60 for the Pro model.
different varieties of apple: Macintosh and Empire. It takes one (a) Find the number of each model that should be produced to
day to trim an acre of Macintosh apples and two days to trim maximize profit and (b) find the maximum profit.
an acre of Empire apples, and there are 240 days per year In Exercises 27–32, the linear programming problem has an
available for trimming. It takes 0.3 day to pick an acre of unusual characteristic. Sketch a graph of the solution region for the
Macintosh apples and 0.1 day to pick an acre of Empire apples, problem and describe the unusual characteristic. (In each problem,
and there are 30 days per year available for picking. Find the objective function is to be maximized.)
the number of acres of each fruit that should be planted to
maximize total profit, assuming that the profits are $140 per 27. Objective function: 28. Objective function:
acre for Macintosh apples and $235 per acre for Empire apples. z ⫽ 2.5x ⫹ y z⫽x⫹y
23. A farming cooperative mixes two brands of cattle feed. Constraints: Constraints:
A standard brand costs $25 per bag and contains 2 units of x ⱖ 0 x ⱖ 0
nutritional element A, 2 units of element B, and 2 units of
y ⱖ 0 y ⱖ 0
element C. An economy brand costs $20 per bag and contains
1 unit of nutritional element A, 9 units of element B, and 3x ⫹ 5y ⱕ 15 ⫺x ⫹ y ⱕ 1
3 units of element C. Find the number of bags of each brand 5x ⫹ 2y ⱕ 10 ⫺x ⫹ 2y ⱕ 4
that should be mixed to produce nine bags of a mixture having 29. Objective function: 30. Objective function:
a minimum cost per bag. The minimum requirements for
z ⫽ ⫺x ⫹ 2y z⫽x⫹y
nutrients A, B, and C are 12 units, 36 units, and 24 units,
respectively. Constraints: Constraints:
x ⱖ 0 x ⱖ 0
24. Two gasolines, regular and premium, have octane ratings of 87
and 92, respectively. Regular costs $.83 per liter and premium y ⱖ 0 y ⱖ 0
costs $.98 per liter. Determine the blend of minimum cost with x ⱕ 10 ⫺x ⫹ y ⱕ 0
an octane rating of at least 90. [Hint: Let x be the fraction of x⫹y ⱕ 7 ⫺3x ⫹ y ⱖ 3
each liter that is regular and y be the fraction that is premium.]
25. An electronics company produces two models of headphones
z ⫽ 3x ⫹ 4y z ⫽ x ⫹ 2y
or ear buds, a standard model and a high-impact water-resistant
model. Suppose the standard model requires 10 minutes to Constraints: Constraints:
manufacture and 5 minutes to assemble and test, while the x ⱖ 0 x ⱖ 0
high-impact model requires 12 minutes to manufacture and y ⱖ 0 y ⱖ 0
18 minutes to assemble and test. The profits for the two models x⫹y ⱕ 1 x ⫹ 2y ⱕ 4
are $45 for the standard model and $65 dollars for the high- 2x ⫹ y ⱖ 4 2x ⫹ y ⱕ 4
impact water-resistant model. The company has 960 minutes of
manufacturing time and 1080 minutes of assembly/testing time
per day. (a) Find the number of each model that the company
should produce per day to maximize profit and (b) find the
maximum profit.
332600_09_2.qxp 4/22/08 4:51 PM Page 548
In Exercises 33–36, determine, for each given vertex, t-values such In Exercises 37–40, find an objective function that has a maximum
that the objective function has a maximum value at the vertex. or minimum value at the indicated vertex of the constraint region
33. Objective function: 34. Objective function: shown below. (There are many correct answers.)
z ⫽ x ⫹ ty z ⫽ 3x ⫹ ty y
Constraints: Constraints:
6
x ≥ 0 x ≥ 0
5
y ≥ 0 y ≥ 0 A(0, 4)
x ≤ 1 x ⫹ 2y ≤ 4 B(4, 3)
3
y ≤ 1 x⫺ y ≤ 1
2
(a) 共0, 0兲 (b) 共1, 0兲 (a) 共0, 0兲 (b) 共1, 0兲 1
C(5, 0)
(c) 共1, 1兲 (d) 共0, 1兲 (c) 共2, 1兲 (d) 共0, 2兲 x
1 2 3 4 6
z ⫽ 3x ⫹ ty z ⫽ x ⫹ ty 37. The maximum occurs at vertex A.
Constraints: Constraints: 38. The maximum occurs at vertex B.
x ≥ 0 x ≥ 0 39. The maximum occurs at vertex C.
y ≥ 0 y ≥ 0 40. The minimum occurs at vertex C.
x ⫹ 3y ≤ 15 x ⫹ 2y ≤ 4 True or False? In Exercises 41 and 42, determine whether each
4x ⫹ y ≤ 16 x⫺ y ≤ 1 statement is true or false. If a statement is true, give a reason or
cite an appropriate statement from the text. If a statement is false,
(a) 共0, 5兲 (b) 共3, 4兲 (a) 共2, 1兲 (b) 共0, 2兲
provide an example that shows the statement is not true in all cases
(c) 共4, 0兲 (d) 共0, 0兲 (c) 共1, 0兲 (d) 共0, 0兲 or cite an appropriate statement from the text.
41. If an objective function has a maximum value at the vertices

共4, 7兲 and 共8, 3兲, you can conclude that it also has a maximum
value at the points 共4.5, 6.5兲 and 共7.8, 3.2兲.
42. When solving a linear programming problem, if the objective
function has a maximum value at more than one vertex, you can
assume that there are an infinite number of points that will
produce the maximum value.
332600_09_3.qxp 4/22/08 4:52 PM Page 550
9.3 The Simplex Method: Maximization

For linear programming problems involving two variables, the graphical solution method
introduced in Section 9.2 is convenient. For problems involving more than two variables or
problems involving large numbers of constraints, it is better to use solution methods that are
adaptable to computers. One such method is called the simplex method, developed by
George Dantzig in 1946. It provides a systematic way of examining the vertices of the
feasible region to determine the optimal value of the objective function. This method is
introduced with the following example.
Suppose you want to find the maximum value of z ⫽ 4x1 ⫹ 6x2, where x1 ⱖ 0 and
x2 ⱖ 0, subject to the following constraints.
⫺x1 ⫹ x2 ⱕ 11
x1 ⫹ x2 ⱕ 27
2x1 ⫹ 5x2 ⱕ 90
Because the left-hand side of each inequality is less than or equal to the right-hand side,
there must exist nonnegative numbers s1, s2, and s3 that can be added to the left side of each
equation to produce the system of linear equations shown below.
⫺x1 ⫹ x2 ⫹ s1 ⫽ 11
x1 ⫹ x2 ⫹ s2 ⫽ 27
2x1 ⫹ 5x2 ⫹ s3 ⫽ 90
The numbers s1, s2, and s3 are called slack variables because they represent the “slack” in
each inequality.
Standard Form of a A linear programming problem is in standard form if it seeks to maximize the objective
Linear Programming function z ⫽ c1x1 ⫹ c2 x2 ⫹ . . . ⫹ cn xn subject to the constraints
Problem a11x1 ⫹ a12 x2 ⫹ . . . ⫹ a1n xn ⱕ b1
a21 x1 ⫹ a22 x2 ⫹ . . . ⫹ a2n xn ⱕ b2
..
.
am1 x1 ⫹ am2 x2 ⫹ . . . ⫹ amn xn ⱕ bm
where xi ⱖ 0 and bi ⱖ 0. After adding slack variables, the corresponding system of
constraint equations is
a11x1 ⫹ a12 x2 ⫹ . . . ⫹ a1n xn ⫹ s1 ⫽ b1
a21x1 ⫹ a22 x2 ⫹ . . . ⫹ a2n xn ⫹ s2 ⫽ b2
..
.
am1x1 ⫹ am2 x2 ⫹ . . . ⫹ amn xn ⫹ sm ⫽ bm
where si ⱖ 0.
332600_09_3.qxp 4/22/08 4:52 PM Page 551
Section 9.3 The Simplex Method: Maximization 551
R E M A R K : Note that for a linear programming problem in standard form, the objective
function is to be maximized, not minimized. (Minimization problems will be discussed in
Sections 9.4 and 9.5.)
A basic solution of a linear programming problem in standard form is a solution

共x1, x2, . . . , xn, s1, s2, . . . , sm兲 of the constraint equations in which at most m variables are
nonzero, and the variables that are nonzero are called basic variables. A basic solution for
which all variables are nonnegative is called a basic feasible solution.
The Simplex Tableau

The simplex method is carried out by performing elementary row operations on a matrix
called the simplex tableau. This tableau consists of the augmented matrix corresponding to
the constraint equations together with the coefficients of the objective function written in
the form
⫺c1x1 ⫺ c2x2 ⫺ . . . ⫺ cnxn ⫹ 共0兲s1 ⫹ 共0兲s2 ⫹ . . . ⫹ 共0兲sm ⫹ z ⫽ 0.
In the tableau, it is customary to omit the coefficient of z. For instance, the simplex tableau
for the linear programming problem
z ⫽ 4x1 ⫹ 6x2 Objective function
⫺x1 ⫹ x2 ⫹ s1 ⫽ 11
x1 ⫹ x2 ⫹ s2 ⫽ 27 Constraints
2x1 ⫹ 5x2 ⫹ s3 ⫽ 90
is as follows.
Basic
x1 x2 s1 s2 s3 b Variables
⫺1 1 1 0 0 11 s1
1 1 0 1 0 27 s2
2 5 0 0 1 90 s3
⫺4 ⫺6 0 0 0 0
→
Current z-value
For this initial simplex tableau, the basic variables are s1, s2, and s3, and the nonbasic
variables are x1 and x2. The nonbasic variables have a value of zero, yielding a current
z-value of zero. From the columns that are farthest to the right, you can see that the basic
variables have initial values of s1 ⫽ 11, s2 ⫽ 27, and s3 ⫽ 90. So the current solution is
x1 ⫽ 0, x2 ⫽ 0, s1 ⫽ 11, s2 ⫽ 27, and s3 ⫽ 90.
332600_09_3.qxp 4/22/08 4:52 PM Page 552
This solution is a basic feasible solution and is often written as

共x1, x2, s1, s2, s3兲 ⫽ 共0, 0, 11, 27, 90兲.
The entry in the lower right corner of the simplex tableau is the current value of z. Note that
the bottom-row entries under x1 and x2 are the negatives of the coefficients of x1 and x2 in
the objective function
z ⫽ 4x1 ⫹ 6x2.
To perform an optimality check for a solution represented by a simplex tableau, look at
the entries in the bottom row of the tableau. If any of these entries are negative (as above),
then the current solution is not optimal.
Pivoting
Once you have set up the initial simplex tableau for a linear programming problem, the
simplex method consists of checking for optimality and then, if the current solution is not
optimal, improving the current solution. (An improved solution is one that has a larger
z-value than the current solution.) To improve the current solution, bring a new basic
variable into the solution, the entering variable. This implies that one of the current basic
variables (the departing variable) must leave, otherwise you would have too many
variables for a basic solution. You choose the entering and departing variables as follows.
1. The entering variable corresponds to the smallest (the most negative) entry in the
bottom row of the tableau.
2. The departing variable corresponds to the smallest nonnegative ratio of bi兾aij in the
column determined by the entering variable.
3. The entry in the simplex tableau in the entering variable’s column and the departing
variable’s row is called the pivot.
Finally, to form the improved solution, apply Gauss-Jordan elimination to the column that
contains the pivot, as illustrated in Example 1. (This process is called pivoting.)
EXAMPLE 1 Pivoting to Find an Improved Solution

Use the simplex method to find an improved solution for the linear programming problem
represented by the tableau shown below.
Basic
⫺1 1 1 0 0 11 s1
1 1 0 1 0 27 s2
2 5 0 0 1 90 s3
⫺4 ⫺6 0 0 0 0
The objective function for this problem is z ⫽ 4x1 ⫹ 6x2.

332600_09_3.qxp 4/22/08 4:52 PM Page 553
SOLUTION Note that the current solution 共x1 ⫽ 0, x2 ⫽ 0, s1 ⫽ 11, s2 ⫽ 27, s3 ⫽ 90兲 corresponds to a
z-value of 0. To improve this solution, you determine that x2 is the entering variable,
because ⫺6 is the smallest entry in the bottom row.
Basic
⫺1 1 1 0 0 11 s1
1 1 0 1 0 27 s2
2 5 0 0 1 90 s3
⫺4 ⫺6 0 0 0 0
→
Entering
To see why you can choose x2 as the entering variable, remember that z ⫽ 4x1 ⫹ 6x2. So,
it appears that a unit change in x2 produces a change of 6 in z, whereas a unit change in x1
produces a change of only 4 in z.
To find the departing variable, locate the bi’s that have corresponding positive elements in
the entering variables column and form the ratios
11 27 90
⫽ 11, ⫽ 27, and ⫽ 18.
1 1 5
Here the smallest positive ratio is 11, so you can choose s1 as the departing variable.
Basic
⫺1 1 1 0 0 11 s1 ← Departing
1 1 0 1 0 27 s2
2 5 0 0 1 90 s3
⫺4 ⫺6 0 0 0 0
→
Entering
Note that the pivot is the entry in the first row and second column. Now, use Gauss-Jordan
elimination to obtain the improved solution shown below.
Before Pivoting After Pivoting
⫺1 ⫺1
冤冥冤冥
1 1 0 0 11 1 1 0 0 11
1 1 0 1 0 27 2 0 ⫺1 1 0 16 ⴚR1 ⴙ R2
2 5 0 0 1 90 7 0 ⫺5 0 1 35 ⴚ5R1 ⴙ R3
⫺4 ⫺6 0 0 0 0 ⫺10 0 6 0 0 66 6R1 ⴙ R4
The new tableau now appears as follows.

332600_09_3.qxp 4/22/08 4:52 PM Page 554
Basic
⫺1 1 1 0 0 11 x2
2 0 ⫺1 1 0 16 s2
7 0 ⫺5 0 1 35 s3
⫺10 0 6 0 0 66
Note that x2 has replaced s1 in the basic variables column and the improved solution
共x1, x2, s1, s2, s3兲 ⫽ 共0, 11, 0, 16, 35兲
has a z-value of
z ⫽ 4x1 ⫹ 6x2 ⫽ 4共0兲 ⫹ 6共11兲 ⫽ 66.
In Example 1, the improved solution is not yet optimal because the bottom row still has
a negative entry. So, you can apply another iteration of the simplex method to improve the
solution further, as follows. Choose x1 as the entering variable. Moreover, the smallest
nonnegative ratio among 11兾共⫺1兲 ⫽ ⫺11, 16兾2 ⫽ 8, and 35兾7 ⫽ 5 is 5, so s3 is the
departing variable. Gauss-Jordan elimination produces the following matrices.
⫺1 ⫺1
冤冥冤冥
1 1 0 0 11 1 1 0 0 11
2 0 ⫺1 1 0 16 2 0 ⫺1 1 0 16
0 ⫺ 57 1
7 0 ⫺5 0 1 35 1 0 7 5 1
7 R3
⫺10 0 6 0 0 66 ⫺10 0 6 0 0 66
2 1
冤冥
0 1 7 0 7 16 R1 ⴙ R3
3
0 0 7 1 ⫺ 27 6 R2 ⴚ 2R3
1 0 ⫺ 57 0 1
7 5
0 0 ⫺ 87 0 10
7 116 R4 ⴙ 10R3
So, the new simplex tableau is as follows.

Basic
2 1
0 1 7 0 7 16 x2
3
0 0 7 1 ⫺ 27 6 s2
1 0 ⫺ 57 0 1
7 5 x1
8 10
0 0 ⫺7 0 7 116
In this tableau, there is still a negative entry in the bottom row. So, choose s1 as the
entering variable and s2 as the departing variable, as shown in the next tableau.
332600_09_3.qxp 4/22/08 4:52 PM Page 555
Basic
2 1
0 1 7 0 7 16 x2
3
0 0 7 1 ⫺ 27 6 s2 ← Departing
⫺ 57 1
1 0 0 7 5 x1
⫺ 87 10
0 0 0 7 116
→
Entering
By performing one more iteration of the simplex method, you obtain the tableau shown
below. (Try checking this.)
Basic
⫺ 23 1
0 1 0 3 12 x2
7
0 0 1 3 ⫺ 23 14 s1
5
1 0 0 3 ⫺ 13 15 x1
8 2
0 0 0 3 3 132 ← Maximum z-value
In this tableau, there are no negative elements in the bottom row. So, the optimal
solution is determined to be
共x1, x2, s1, s2, s3兲 ⫽ 共15, 12, 14, 0, 0兲
with
z ⫽ 4x1 ⫹ 6x2 ⫽ 4共15兲 ⫹ 6共12兲 ⫽ 132.
R E M A R K : Ties may occur in choosing entering and/or departing variables. Should this
happen, any choice among the tied variables may be made.
x2 Because the linear programming problem in Example 1 involved only two decision
25
variables, you could have used a graphical solution technique, as in Example 2, Section 9.2.
Notice in Figure 9.18 that each iteration in the simplex method corresponds to moving from
20 a given vertex to an adjacent vertex with an improved z-value.
(5, 16)
15 (15, 12) 共0, 0兲共0, 11兲共5, 16兲共15, 12兲
10 (0, 11)
z⫽0 z ⫽ 66 z ⫽ 116 z ⫽ 132
5
(0, 0) (27, 0)
x1
5 10 15 20 25 30 The Simplex Method
Figure 9.18
The steps involved in the simplex method can be summarized as follows.
332600_09_3.qxp 4/22/08 4:52 PM Page 556
The Simplex Method To solve a linear programming problem in standard form, use the following steps.
(Standard Form) 1. Convert each inequality in the set of constraints to an equation by adding slack
variables.
2. Create the initial simplex tableau.
3. Locate the most negative entry in the bottom row. The column for this entry is called
the entering column. (If ties occur, any of the tied entries can be used to determine the
entering column.)
4. Form the ratios of the entries in the “b-column” with their corresponding positive
entries in the entering column. The departing row corresponds to the smallest nonneg-
ative ratio bi /aij. (If all entries in the entering column are 0 or negative, then there is no
maximum solution. For ties, choose either entry.) The entry in the departing row and
the entering column is called the pivot.
5. Use elementary row operations so that the pivot is 1, and all other entries in the
entering column are 0. This process is called pivoting.
6. If all entries in the bottom row are zero or positive, this is the final tableau. If not, go
back to step 3.
7. If you obtain a final tableau, then the linear programming problem has a maximum
solution, which is given by the entry in the lower right corner of the tableau.
Note that the basic feasible solution of an initial simplex tableau is

共x1, x2, . . . , xn, s1, s2, . . . , sm 兲 ⫽ 共0, 0, . . . , 0, b1, b2, . . . , bm 兲.
This solution is basic because at most m variables are nonzero (namely the slack variables).
It is feasible because each variable is nonnegative.
In the next two examples, the use of the simplex method to solve a problem involving
three decision variables is illustrated.
EXAMPLE 2 The Simplex Method with Three Decision Variables

Use the simplex method to find the maximum value of
z ⫽ 2x1 ⫺ x2 ⫹ 2x3 Objective function
subject to the constraints

2x1 ⫹ x2 ⱕ 10
x1 ⫹ 2x2 ⫺ 2x3 ⱕ 20
x2 ⫹ 2x3 ⱕ 5
where x1 ⱖ 0, x2 ⱖ 0, and x3 ⱖ 0.
SOLUTION Using the basic feasible solution,
共x1, x2, x3, s1, s2, s3兲 ⫽ 共0, 0, 0, 10, 20, 5兲,
332600_09_3.qxp 4/22/08 4:52 PM Page 557
the initial simplex tableau for this problem is as follows. (Try checking these computations,
and note the “tie” that occurs when choosing the first entering variable.)
Basic
x1 x2 x3 s1 s2 s3 b Variables
2 1 0 1 0 0 10 s1
1 2 ⫺2 0 1 0 20 s2
0 1 2 0 0 1 5 s3 ← Departing
⫺2 1 ⫺2 0 0 0 0
→
Entering
Basic
2 1 0 1 0 0 10 s1 ← Departing
1 3 0 0 1 1 25 s2
1 1 5
0 2 1 0 0 2 2 x3
⫺2 2 0 0 0 1 5
→
Entering
Basic
1 1
1 2 0 2 0 0 5 x1
5
0 2 0 ⫺ 12 1 1 20 s2
1 1 5
0 2 1 0 0 2 2 x3
0 3 0 1 0 1 15
This implies that the optimal solution is
共x1, x2, x3, s1, s2, s3兲 ⫽ 共5, 0, 52, 0, 20, 0兲
and the maximum value of z is 15.
Note that s2 ⫽ 20. The optimal solution yields a maximum value of z ⫽ 15 provided that
x1 ⫽ 5, x2 ⫽ 0, and x3 ⫽ 52. Notice that these values satisfy the constraints giving equality
in the first and third constraints, yet the second constraint has a slack of 20.
Occasionally, the constraints in a linear programming problem will include an equation.

In such cases, you can still add a “slack variable” called an artificial variable to form the
initial simplex tableau. Technically, this new variable is not a slack variable (because there
is no slack to be taken). Once you have determined an optimal solution in such a problem,
you should check to see that any equations in the original constraints are satisfied. Example
3 illustrates such a case.
332600_09_3.qxp 4/22/08 4:52 PM Page 558
EXAMPLE 3 The Simplex Method with Three Decision Variables

Use the simplex method to find the maximum value of
z ⫽ 3x1 ⫹ 2x2 ⫹ x3 Objective function

4x1 ⫹ x2 ⫹ x3 ⫽ 30
2x1 ⫹ 3x2 ⫹ x3 ⱕ 60
x1 ⫹ 2x2 ⫹ 3x3 ⱕ 40
SOLUTION Using the basic feasible solution,
共x1, x2, x3, s1, s2, s3兲 ⫽ 共0, 0, 0, 30, 60, 40兲,
the initial simplex tableau for this problem is as follows. (Note that s1 is an artificial
variable, rather than a slack variable.)
Basic
4 1 1 1 0 0 30 s1 ← Departing
2 3 1 0 1 0 60 s2
1 2 3 0 0 1 40 s3
⫺3 ⫺2 ⫺1 0 0 0 0
→
Entering
Basic
1 1 1 15
1 4 4 4 0 0 2 x1
5 1
0 2 2 ⫺ 12 1 0 45 s2 ← Departing
7 11
⫺ 14 65
0 4 4 0 1 2 s3
⫺ 54 ⫺ 14 3 45
0 4 0 0 2
→
Entering
Basic
1 3 1
1 0 5 10 ⫺ 10 0 3 x1
1
⫺ 15 2
0 1 5 5 0 18 x2
12 1 7
0 0 5 10 ⫺ 10 1 1 s3
1 1
0 0 0 2 2 0 45
332600_09_3.qxp 4/22/08 4:52 PM Page 559
This implies that the optimal solution is

共x1, x2, x3, s1, s2, s3兲 ⫽ 共3, 18, 0, 0, 0, 1兲
and the maximum value of z is 45. (This solution satisfies the equation provided in the
constraints, because 4共3兲 ⫹ 1共18兲 ⫹ 1共0兲 ⫽ 30.)
Applications
EXAMPLE 4 A Business Application: Maximum Profit

A manufacturer produces three types of plastic fixtures. The times required for molding,
trimming, and packaging are presented in Table 9.1. (Times and profits are displayed in
hours per dozen fixtures.)
TABLE 9.1
Process Type A Type B Type C Total time available

3
Molding 1 2 2 12,000
2 2
Trimming 3 3 1 4600
1 1 1
Packaging 2 3 2 2400
Profit $11 $16 $15 —
How many dozen of each type of fixture should be produced to obtain a maximum profit?
SOLUTION Letting x1, x2, and x3 represent the numbers of dozens of types A, B, and C fixtures,
respectively, the objective function is expressed as
Profit ⫽ P ⫽ 11x1 ⫹ 16x2 ⫹ 15x3.
Moreover, using the information in the table, you can construct the constraints listed below.
x1 ⫹ 2x2 ⫹ 32 x3 ⱕ 12,000
2 2
3 x1 ⫹ 3 x2 ⫹ x3 ⱕ 4600
1 1 1
2 x1 ⫹ 3 x2 ⫹ 2 x3 ⱕ 2400
共Also assume that x1 ⱖ 0, x2 ⱖ 0, and x3 ⱖ 0.兲 Now, apply the simplex method with the
basic feasible solution
共x1, x2, x3, s1, s2, s3兲 ⫽ 共0, 0, 0, 12,000, 4600, 2400兲
to obtain the following tableau.
332600_09_3.qxp 4/22/08 4:52 PM Page 560
Basic
3
1 2 2 1 0 0 12,000 s1 ← Departing
2 2
3 3 1 0 1 0 4600 s2
1 1 1
2 3 2 0 0 1 2400 s3
⫺11 ⫺16 ⫺15 0 0 0 0
→
Entering
Basic
1 3 1
2 1 4 2 0 0 6000 x2
1 1
3 0 2 ⫺ 13 1 0 600 s2
1 1
3 0 4 ⫺ 16 0 1 400 s3 ← Departing
⫺3 0 −3 8 0 0 96,000
→
Entering
Basic
3 3
0 1 8 4 0 ⫺ 32 5400 x2
1
0 0 4 ⫺ 16 1 ⫺1 200 s2 ← Departing
3
1 0 4 ⫺ 12 0 3 1200 x1
⫺ 34 13
0 0 2 0 9 99,600
→
Entering
Basic
0 1 0 1 ⫺ 32 0 5100 x2
0 0 1 ⫺ 23 4 ⫺4 800 x3
1 0 0 0 ⫺3 6 600 x1
0 0 0 6 3 6 100,200
From this final simplex tableau, you can see that the maximum profit is $100,200, and this
is obtained by using the production levels listed below.
Type A: 600 dozen units
Type B: 5100 dozen units
Type C: 800 dozen units
332600_09_3.qxp 4/22/08 4:52 PM Page 561
R E M A R K : In Example 4, note that the second simplex tableau contains a “tie” for the
minimum entry in the bottom row. 共Both the first and third entries in the bottom row are
⫺3.兲 Although the first column was chosen to represent the departing variable, the third
column could have also been chosen. Try reworking the problem with this choice to see that
you obtain the same solution.
EXAMPLE 5 A Business Application: Media Selection

The advertising alternatives for a company include television, radio, and newspaper
advertisements. The costs and estimates of audience coverage are provided in Table 9.2.
TABLE 9.2
Television Newspaper Radio
Cost per advertisement $2000 $600 $300
Audience per advertisement 100,000 40,000 18,000
The local newspaper limits the number of weekly advertisements from a single company to
ten. Moreover, in order to balance the advertising among the three types of media, no more
than half of the total number of advertisements should occur on the radio, and at least 10%
should occur on television. The weekly advertising budget is $18,200. How many advertise-
ments should be run in each of the three types of media to maximize the total audience?
SOLUTION To begin, let x1, x2, and x3 represent the numbers of advertisements in television, newspaper,
and radio, respectively. The objective function (to be maximized) is
z ⫽ 100,000x1 ⫹ 40,000x2 ⫹ 18,000x3 Objective function
where x1 ⱖ 0, x2 ⱖ 0, and x3 ⱖ 0. The constraints for this problem are as follows.

2000x1 ⫹ 600x2 ⫹ 300x3 ⱕ 18,200
x2 ⱕ 10
x3 ⱕ 0.5 共x1 ⫹ x2 ⫹ x3兲
x1 ⱖ 0.1共x1 ⫹ x2 ⫹ x3兲
A more manageable form of this system of constraints is as follows.
20x1 ⫹ 6x2 ⫹ 3x3 ⱕ 182
x2 ⱕ 10
Constraints
⫺x1 ⫺ x2 ⫹ x3 ⱕ 0
⫺9x1 ⫹ x2 ⫹ x3 ⱕ 0
So, the initial simplex tableau is as follows.
332600_09_3.qxp 4/22/08 4:52 PM Page 562
Basic
x1 x2 x3 s1 s2 s3 s4 b Variables
20 6 3 1 0 0 0 182 s1 ← Departing
0 1 0 0 1 0 0 10 s2
⫺1 ⫺1 1 0 0 1 0 0 s3
⫺9 1 1 0 0 0 1 0 s4
⫺100,000 ⫺40,000 ⫺18,000 0 0 0 0 0
→
Entering
Now, to this initial tableau, apply the simplex method as follows.

Basic
3 3 1 91
1 10 20 20 0 0 0 10 x1
0 1 0 0 1 0 0 10 s2 ← Departing
7 23 1 91
0 ⫺ 10 20 20 0 1 0 10 s3
37 47 9 819
0 10 20 20 0 0 1 10 s4
0 ⫺10,000 ⫺3000 5000 0 0 0 910,000

→
Entering
Basic
3 1 3 61
1 0 20 20 ⫺ 10 0 0 10 x1
0 1 0 0 1 0 0 10 x2
23 1 7 161
0 0 20 20 10 1 0 10 s3 ← Departing
47 9
⫺ 37 449
0 0 20 20 10 0 1 10 s4
0 0 ⫺3000 5000 10,000 0 0 1,010,000

→
Entering
Basic
1 9 3
1 0 0 23 ⫺ 23 ⫺ 23 0 4 x1
0 1 0 0 1 0 0 10 x2
1 14 20
0 0 1 23 23 23 0 14 x3
8
0 0 0 23 ⫺ 118
23 ⫺ 47
23 1 12 s4
118,000 272,000 60,000
0 0 0 23 23 23 0 1,052,000
332600_09_3.qxp 4/22/08 4:52 PM Page 563
From the final tableau on the preceding page, you can see that the maximum weekly
audience for an advertising budget of $18,200 is
z ⫽ 1,052,000 Maximum weekly audience
and this occurs when x1 ⫽ 4, x2 ⫽ 10, and x3 ⫽ 14. The results are summarized as follows.
Number of
Media Advertisements Cost Audience
Television 4 $8000 400,000

Newspaper 10 $6000 400,000
Radio 14 $4200 252,000
Total 28 $18,200 1,052,000

In Exercises 1–4, write the simplex tableau for the linear In Exercises 5–8, explain why the linear programming problem is
programming problem. You do not need to solve the problem. (In not in standard form.
each case, the objective function is to be maximized.) 5. (Minimize) 6. (Maximize)
1. Objective function: 2. Objective function: Objective function: Objective function:
z ⫽ x1 ⫹ 2x2 z ⫽ x1 ⫹ 3x2 z ⫽ x1 ⫹ x2 z ⫽ x1 ⫹ x2
Constraints: Constraints: Constraints: Constraints:
2x1 ⫹ x2 ⱕ 8 x1 ⫹ x2 ⱕ 4 x1 ⫹ 2x2 ⱕ 4 x1 ⫹ 2x2 ⱕ 6
x1 ⫹ x2 ⱕ 5 x1 ⫺ x2 ⱕ 1 x1, x2, ⱖ 0 2x1 ⫺ x2 ⱕ ⫺1
x1, x2 ⱖ 0 x1, x2 ⱖ 0 x1, x2 ⱖ 0
7. (Maximize) 8. (Maximize)
Objective function: Objective function:
z ⫽ 2x1 ⫹ 3x2 ⫹ 4x3 z ⫽ 6x1 ⫺ 9x2
z ⫽ x1 ⫹ x2 z ⫽ x1 ⫹ x2
x1 ⫹ 2x2 ⱕ 12 2x1 ⫺ 3x2 ⱕ 6
x1 ⫹ x2 ⫹ 3x3 ⱕ 5 x1 ⫹ x2 ⱖ 4
x1 ⫹ x3 ⱕ 8 x1 ⫹ x2 ⱕ 20
2x1 ⫺ 2x3 ⱖ 1 2x1 ⫹ x2 ⱖ 6
x1, x2, x3 ⱖ 0 x1, x2 ⱖ 0
x2 ⫹ x3 ⱕ 0 x1, x2 ⱖ 0
x1, x2, x3 ⱖ 0
332600_09_3.qxp 4/22/08 4:52 PM Page 564
In Exercises 9–24, use the simplex method to solve the linear 19. Objective function: 20. Objective function:
programming problem. (In each case, the objective function is to be z ⫽ 3x1 ⫹ 4x2 ⫹ x3 ⫹ 7x4 z ⫽ x1
maximized.)
8x1 ⫹ 3x2 ⫹ 4x3 ⫹ x4 ⱕ 7 3x1 ⫹ 2x2 ⱕ 60
z ⫽ x1 ⫹ 2x2 z ⫽ x1 ⫹ x2
2x1 ⫹ 6x2 ⫹ x3 ⫹ 5x4 ⱕ 3 x1 ⫹ 2x2 ⱕ 28
x1 ⫹ 4x2 ⫹ 5x3 ⫹ 2x4 ⱕ 8 x1 ⫹ 4x2 ⱕ 48
x1 ⫹ 4x2 ⱕ 8 x1 ⫹ 2x2 ⱕ 6
x1, x2, x3, x4 ⱖ 0 x1, x2 ⱖ 0
x1 ⫹ x2 ⱕ 12 3x1 ⫹ 2x2 ⱕ 12
x1, x2 ⱖ 0 x1, x2 ⱖ 0
z ⫽ x1 ⫺ x2 ⫹ x3 z ⫽ 2x1 ⫹ x2 ⫹ 3x3
z ⫽ 3x1 ⫹ 2x2 z ⫽ 4x1 ⫹ 5x2
2x1 ⫹ x2 ⫺ 3x3 ⱕ 40 x1 ⫹ x2 ⫹ x3 ⱕ 59
x1 ⫹ x3 ⱕ 25 2x1 ⫹ 3x3 ⱕ 75
x1 ⫹ 3x2 ⱕ 15 2x1 ⫹ 2x2 ⱕ 10
2x2 ⫹ 3x3 ⱕ 32 x2 ⫹ 6x3 ⱕ 54
4x1 ⫹ x2 ⱕ 16 x1 ⫹ 2x2 ⱕ 6
x1, x2, x3 ⱖ 0 x1, x2, x3 ⱖ 0
x1, x2 ⱖ 0 x1, x2 ⱖ 0
23. Objective function:
z ⫽ x1 ⫹ 2x2 ⫺ x4
z ⫽ 10x1 ⫹ 7x 2 z ⫽ 25x1 ⫹ 35x2
Constraints:
x1 ⫹ 2x2 ⫹ 3x3 ⱕ 24
x1 ⱕ 60 18x1 ⫹ 9x2 ⱕ 7200
3x2 ⫹ 7x3 ⫹ x4 ⱕ 42
x2 ⱕ 45 8x1 ⫹ 9x2 ⱕ 3600
x1, x2, x3, x4 ⱖ 0
5x1 ⫹ 6x2 ⱕ 420 x1, x2 ⱖ 0
24. Objective function:
x1, x2 ⱖ 0
z ⫽ x1 ⫹ 2x2 ⫹ x3 ⫺ x4
Constraints:
z ⫽ 4x1 ⫹ 5x2 z ⫽ x1 ⫹ 2x2
x1 ⫹ x2 ⫹ 3x3 ⫹ 4x4 ⱕ 60
x2 ⫹ 2x3 ⫹ 5x4 ⱕ 50
x1 ⫹ x2 ⱕ 10 x1 ⫹ 3x2 ⱕ 15
2x1 ⫹ 3x2 ⫹ 6x4 ⱕ 72
3x1 ⫹ 7x2 ⱕ 42 2x1 ⫺ x2 ⱕ 12
x1, x2, x3, x4 ⱖ 0
x1, x2 ⱖ 0 x1, x2 ⱖ 0
25. An online electronics website plans to sell two models of home
computers at costs of $250 and $400. The $250 model yields a
z ⫽ 5x1 ⫹ 2x 2 ⫹ 8x3 z ⫽ x1 ⫺ x2 ⫹ 2x3 profit of $45 and the $400 model yields a profit of $50. The
Constraints: Constraints: website estimates that the total monthly demand will not exceed
2x1 ⫺ 4x2 ⫹ x3 ⱕ 42 2x1 ⫹ 2x2 ⱕ 8 250 units. Find the number of units of each model that should be
stocked in order to maximize profit. Assume that the website
2x1 ⫹ 3x2 ⫺ x3 ⱕ 42 x3 ⱕ 5 does not want to invest more than $70,000 in computer
6x1 ⫺ x2 ⫹ 3x3 ⱕ 42 x1, x2, x3 ⱖ 0 inventory. (See Exercise 21 in Section 9.2.)
x1, x2, x3 ⱖ 0
332600_09_3.qxp 4/22/08 4:52 PM Page 565
26. A fruit grower has 150 acres of land available to raise two 31. A company has budgeted a maximum of $600,000 for advertis-
varieties of apples: Macintosh and Empire. It takes one day to ing a certain product nationally. Each minute of television time
trim an acre of Macintosh apples and two days to trim an acre of costs $60,000 and each one-page newspaper ad costs $15,000.
Empire apples, and there are 240 days per year available for Each television ad is expected to be viewed by 15 million view-
trimming. It takes 0.3 day to pick an acre of Macintosh apples ers, and each newspaper ad is expected to be seen by 3 million
and 0.1 day to pick an acre of Empire apples, and there are readers. The company’s market research department advises the
30 days per year available for picking. Find the number of acres company to spend at most 90% of the advertising budget on
of each fruit that should be planted to maximize total profit, television ads. How should the advertising budget be allocated
assuming that the profits are $140 per acre for Macintosh apples to maximize the total audience?
and $235 per acre for Empire apples. (See Exercise 22 in 32. Rework Exercise 31 assuming that each one-page newspaper ad
Section 9.2.) costs $30,000.
27. A grower has 50 acres of land on which she plans to raise three 33. An investor has up to $250,000 to invest in three types of
crops. It costs $200 to produce an acre of carrots and the profit investments. Type A pays 8% annually and has a risk factor of
is $60 per acre. It costs $80 to produce an acre of celery and the 0. Type B pays 10% annually and has a risk factor of 0.06. Type
profit is $20 per acre. Finally, it costs $140 to produce an acre of C pays 14% annually and has a risk factor of 0.10. To have a
lettuce and the profit is $30 per acre. Use the simplex method to well-balanced portfolio, the investor imposes the following
find the number of acres of each crop she should plant in order conditions. The average risk factor should be no greater than
to maximize her profit. Assume her cost cannot exceed $10,000. 0.05. Moreover, at least one-fourth of the total portfolio is to be
28. A fruit juice company makes two special drinks by blending allocated to type A investments and at least one-fourth of the
apple and pineapple juices. The first drink uses 30% apple juice portfolio is to be allocated to type B investments. How much
and 70% pineapple juice, while the second drink uses 60% apple should be allocated to each type of investment to obtain a
juice and 40% pineapple juice. There are 1000 liters of apple maximum return?
juice and 1500 liters of pineapple juice available. If the profit for 34. An investor has up to $450,000 to invest in three types of
the first drink is $.60 per liter and that for the second drink is investments. Type A pays 6% annually and has a risk factor of
$.50, use the simplex method to find the number of liters of each 0. Type B pays 10% annually and has a risk factor of 0.06. Type
drink that should be produced in order to maximize the profit. C pays 12% annually and has a risk factor of 0.08. To have a
29. A manufacturer produces three models of bicycles. The times well-balanced portfolio, the investor imposes the following
(in hours) required for assembling, painting, and packaging each conditions. The average risk factor should be no greater than
model are shown in the table below. 0.05. Moreover, at least one-half of the total portfolio is to be
allocated to type A investments and at least one-fourth of the
Model A Model B Model C portfolio is to be allocated to type B investments. How much
Assembling 2 2.5 3 should be allocated to each type of investment to obtain a
maximum return?
Painting 1.5 2 1 35. An accounting firm has 900 hours of staff time and 100 hours of
Packaging 1 0.75 1.25 reviewing time available each week. The firm charges $2000 for
an audit and $300 for a tax return. Each audit requires 100 hours
The total times available for assembling, painting, and pack- of staff time and 10 hours of review time, and each tax return
aging are 4006 hours, 2495 hours, and 1500 hours, respectively. requires 12.5 hours of staff time and 2.5 hours of review time.
The profits per unit for the three models are $45 (model A), What numbers of audits and tax returns will bring in a maximum
$50 (model B), and $55 (model C). How many of each type revenue?
should be produced to obtain a maximum profit? 36. The accounting firm in Exercise 35 raises its charge for an audit
30. Suppose that in Exercise 29 the total times available for to $2500. What numbers of audits and tax returns will bring in a
assembling, painting, and packaging are 4000 hours, 2500 maximum revenue?
hours, and 1500 hours, respectively, and that the profits per unit
are $48 (model A), $50 (model B), and $52 (model C). How
many of each type should be produced to obtain a maximum
profit?
332600_09_3.qxp 4/22/08 4:52 PM Page 566
In the simplex method, it may happen that in selecting the depart- 41. Use a computer software program to maximize the objective
ing variable all the calculated ratios are negative. This indicates an function
unbounded solution. Demonstrate this in Exercises 37 and 38. z ⫽ 2x1 ⫹ 7x2 ⫹ 6x3 ⫹ 4x4
37. (Maximize) 38. (Maximize) subject to the constraints
x1 ⫹ x2 ⫹ 0.83x3 ⫹ 0.5x4 ⱕ 65
z ⫽ x1 ⫹ 2x2 z ⫽ x1 ⫹ 3x2
1.2x1 ⫹ x2 ⫹ x3 ⫹ 1.2x4 ⱕ 96
0.5x1 ⫹ 0.7x2 ⫹ 1.2x3 ⫹ 0.4x4 ⱕ 80
x1 ⫺ 3x2 ⱕ 1 ⫺x1 ⫹ x2 ⱕ 20
where x1, x2, x3, x4 ⱖ 0.
⫺x1 ⫹ 2x2 ⱕ 4 ⫺2x1 ⫹ x2 ⱕ 50
42. Use a computer software program to maximize the objective
x1, x2 ⱖ 0 x1, x2 ⱖ 0 function
If the simplex method terminates and one or more variables not z ⫽ 1.2x1 ⫹ x2 ⫹ x3 ⫹ x4
in the final basis have bottom-row entries of zero, bringing these subject to the same set of constraints given in Exercise 41.
variables into the basis will determine other optimal solutions.
Demonstrate this in Exercises 39 and 40. True or False? In Exercises 43 and 44, determine whether each
39. (Maximize) 40. (Maximize) statement is true or false. If a statement is true, give a reason or
Objective function: Objective function: cite an appropriate statement from the text. If a statement is false,
1 provide an example that shows the statement is not true in all cases
z ⫽ 2.5x1 ⫹ x2 z ⫽ x1 ⫹ 2x2 or cite an appropriate statement from the text.
43. After creating the initial simplex tableau, the entering column is
3x1 ⫹ 5x2 ⱕ 15 2x1 ⫹ x2 ⱕ 20 chosen by locating the most negative entry in the bottom row.
5x1 ⫹ 2x2 ⱕ 10 x1 ⫹ 3x2 ⱕ 35 44. If all entries in the entering column are 0 or negative, then there
x1, x2 ⱖ 0 x1, x2 ⱖ 0 is no maximum solution.
332600_09_4.qxp 4/22/08 4:52 PM Page 567
Section 9.4 The Simplex Method: Minimization 567
9.4 The Simplex Method: Minimization

In Section 9.3, the simplex method was applied only to linear programming problems
in standard form where the objective function was to be maximized. In this section, this
procedure will be extended to linear programming problems in which the objective function
is to be minimized.
A minimization problem is in standard form if the objective function w ⫽ c1 x1 ⫹ c2 x2
⫹ . . . ⫹ cn xn is to be minimized, subject to the constraints
a11x1 ⫹ a12 x2 ⫹ . . . ⫹ a1n xn ⱖ b1
a21x1 ⫹ a22 x2 ⫹ . . . ⫹ a2n xn ⱖ b2
..
.
am1x1 ⫹ am2 x2 ⫹ . . . ⫹ amn xn ⱖ bm
where xi ⱖ 0 and bi ⱖ 0. The basic procedure used to solve such a problem is to convert
it to a maximization problem in standard form, and then apply the simplex method as
discussed in Section 9.3.
In Example 5 in Section 9.2, geometric methods were used to solve the minimization
problem shown below.
Minimization Problem: Find the minimum value of
w ⫽ 0.12x1 ⫹ 0.15x2 Objective function

60x1 ⫹ 60x2 ⱖ 300
12x1 ⫹ 6x2 ⱖ 36 Constraints
10x1 ⫹ 30x2 ⱖ 90
where x1 ⱖ 0 and x2 ⱖ 0. The first step in converting this problem to a maximization prob-
lem is to form the augmented matrix for this system of inequalities. To this augmented
matrix, add a last row that represents the coefficients of the objective function, as follows.
⯗
冤冥
60 60 300
12 6 ⯗ 36
10 30 ⯗ 90
... ... ... ...
0.12 0.15 ⯗ 0
Next, form the transpose of this matrix by interchanging its rows and columns.
⯗
冤冥
60 12 10 0.12
60 6 30 ⯗ 0.15
... ... ... ... ...
300 36 90 ⯗ 0
332600_09_4.qxp 4/22/08 4:52 PM Page 568
Note that the rows of this matrix are the columns of the first matrix, and vice versa. Finally,
interpret the new matrix as a maximization problem as follows. (To do this, introduce new
variables, y1, y2, and y3.) This corresponding maximization problem is called the dual of the
original minimization problem.
Dual Maximization Problem: Find the maximum value of
z ⫽ 300y1 ⫹ 36y2 ⫹ 90y3 Dual objective function

60y1 ⫹ 12y2 ⫹ 10y3 ⱕ 0.12
Constraints
60y1 ⫹ 6y2 ⫹ 30y3 ⱕ 0.15
where y1 ⱖ 0, y2 ⱖ 0, and y3 ⱖ 0.
As it turns out, the solution of the original minimization problem can be found by
applying the simplex method to the new dual problem, as follows.
Basic
y1 y2 y3 s1 s2 b Variables
60 12 10 1 0 0.12 s1 ← Departing
60 6 30 0 1 0.15 s2
⫺300 ⫺36 ⫺90 0 0 0

→
Entering
Basic
1 1 1 1
1 5 6 60 0 500 y1
3
0 –6 20 –1 1 100 s2 ← Departing
3
0 24 – 40 5 0 5
→
Entering
Basic
1 1 1 7
1 4 0 40 ⫺ 120 4000 y1
3 1 1 3
0 ⫺ 10 1 ⫺ 20 20 2000 y3
33
0 12 0 3 2 50
→
x1 x2
So, the solution of the dual maximization problem is z ⫽ 33 50 ⫽ 0.66. This is the same
value that was obtained in the minimization problem in Example 5 in Section 9.2. The
x-values corresponding to this optimal solution are obtained from the entries in the bottom
row corresponding to slack variable columns. In other words, the optimal solution occurs
when x1 ⫽ 3 and x2 ⫽ 2.
332600_09_4.qxp 4/22/08 4:52 PM Page 569
The fact that a dual maximization problem has the same solution as its original
minimization problem is stated formally in a result called the von Neumann Duality
Principle, after the American mathematician John von Neumann (1903–1957).
THEOREM 9.2 The objective value w of a minimization problem in standard form has a minimum value
The von Neumann if and only if the objective value z of the dual maximization problem has a maximum
Duality Principle value. Moreover, the minimum value of w is equal to the maximum value of z.
Solving a Minimization Problem

The steps used to solve a minimization problem can be summarized, as follows.
Solving a Minimization A minimization problem is in standard form if the objective function w ⫽ c1x1 ⫹ c 2x 2
Problem ⫹ . . . ⫹ cn x n is to be minimized, subject to the constraints
a11x1 ⫹ a12 x2 ⫹ . . . ⫹ a1n xn ⱖ b1
a21 x1 ⫹ a22 x2 ⫹ . . . ⫹ a2n xn ⱖ b2
..
.
am1x1 ⫹ am2 x2 ⫹ . . . ⫹ amn xn ⱖ bm
where xi ⱖ 0 and bi ⱖ 0. To solve this problem, use the following steps.
1. Form the augmented matrix for the given system of inequalities, and add a bottom
row consisting of the coefficients of the objective function.
⯗
冤冥
a11 a12 ... a1n b1
a21 a22 ... a2n ⯗ b2
⯗
am1 am2 ... amn ⯗ bm
... ... ... ... ⯗ ...
c1 c2 ... cn ⯗ 0
2. Form the transpose of this matrix.
⯗
冤冥
a11 a21 ... am1 c1
a12 a22 ... am2 ⯗ c2
⯗
a1n a2n ... amn ⯗ cn
... ... ... ... ⯗ ...
b1 b2 ... bm ⯗ 0
332600_09_4.qxp 4/22/08 4:52 PM Page 570
Solving a Minimization 3. Form the dual maximization problem corresponding to this transposed matrix. That
Problem (continued) is, find the maximum of the objective function given by z ⫽ b1y1 ⫹ b2 y2 ⫹ . . .
⫹ bm ym subject to the constraints
a11 y1 ⫹ a21 y2 ⫹ . . . ⫹ am1 ym ⱕ c1
a12 y1 ⫹ a22 y2 ⫹ . . . ⫹ am2 ym ⱕ c2
..
.
a1n y1 ⫹ a2n y2 ⫹ . . . ⫹ amn ym ⱕ cn
where y1 ⱖ 0, y2 ⱖ 0, . . . , and ym ⱖ 0.
4. Apply the simplex method to the dual maximization problem. The maximum value of
z will be the minimum value of w. Moreover, the values of x1, x2, . . . , xn will occur in
the bottom row of the final simplex tableau, in the columns corresponding to the slack
variables.
The steps used to solve a minimization problem are illustrated in Examples 1 and 2.
EXAMPLE 1 Solving a Minimization Problem

Find the minimum value of
w ⫽ 3x1 ⫹ 2x2 Objective function

2x1 ⫹ x2 ⱖ 6
Constraints
x1 ⫹ x2 ⱖ 4
where x1 ⱖ 0 and x2 ⱖ 0.
SOLUTION The augmented matrix corresponding to this minimization problem is
⯗
冤冥
2 1 6
1 1 ⯗ 4
... ... ⯗ ... .
3 2 ⯗ 0
So, the matrix corresponding to the dual maximization problem is shown by the transpose
冤冥
2 1 3
1 1 ⯗ 2
... ... ⯗ ... .
6 4 ⯗ 0
This implies that the dual maximization problem is as follows.
332600_09_4.qxp 4/22/08 4:52 PM Page 571

z ⫽ 6y1 ⫹ 4y2 Dual objective function

2y1 ⫹ y2 ⱕ 3
Constraints
y1 ⫹ y2 ⱕ 2
where y1 ⱖ 0 and y2 ⱖ 0. Now apply the simplex method to the dual problem, as follows.
Basic
y1 y2 s1 s2 b Variables
2 1 1 0 3 s1 ← Departing
1 1 0 1 2 s2
⫺6 ⫺4 0 0 0
→
Entering
Basic
y1 y2 s1 s2 b Variables
1 1 3
1 2 2 0 2 y1
1
⫺ 12 1
0 2 1 2 s2 ← Departing
y2
(a) 0 ⫺1 3 0 9
→
3 Maximum:
Entering
z = 6y1 + 4y2 = 10
Basic
(0, 2) y1 y2 s1 s2 b Variables
1 (1, 1) 1 0 1 ⫺1 1 y1
0 1 ⫺1 2 1 y2
y1 0 0 2 2 10
(0, 0) ( 32, 0( 2 3
→
→
x1 x2
x2
(b) From this final simplex tableau, you can see that the maximum value of z is 10. So, the
6 (0, 6) solution of the original minimization problem is
Minimum:
5
w = 3x1 + 2x 2 = 10 w ⫽ 10, Minimum value
3 and this occurs when

2 (2, 2) x1 ⫽ 2 and x2 ⫽ 2.
1
(4, 0)
x1
1 2 3 4 5 6 Both the minimization and maximization linear programming problems in Example 1
could have been solved with a graphical method, as indicated in Figure 9.19. Note in
Figure 9.19
Figure 9.19 that the maximum value of z ⫽ 6y1 ⫺ 4y2 in part (a) is the same as the
minimum value of w ⫽ 3x1 ⫹ 2x2 in part (b).
332600_09_4.qxp 4/22/08 4:52 PM Page 572
EXAMPLE 2 Solving a Minimization Problem

w ⫽ 2x1 ⫹ 10x2 ⫹ 8x3 Objective function

x1 ⫹ x2 ⫹ x3 ⱖ 6
x2 ⫹ 2x3 ⱖ 8 Constraints
⫺x1 ⫹ 2x2 ⫹ 2x3 ⱖ 4
SOLUTION The augmented matrix corresponding to this minimization problem is
⯗
冤冥
1 1 1 6
0 1 2 ⯗ 8
⫺1 2 2 ⯗ 4 .
... ... ... ⯗ ...
2 10 8 ⯗ 0
So, the matrix corresponding to the dual maximization problem is shown by the transpose
⯗
冤冥
1 0 ⫺1 2
1 1 2 ⯗ 10
1 2 2 ⯗ 8 .
... ... ... ⯗ ...
6 8 4 ⯗ 0
This implies that the dual maximization problem is as follows.
z ⫽ 6y1 ⫹ 8y2 ⫹ 4y3 Dual objective function

y1 ⫺ y3 ⱕ 2
y1 ⫹ y2 ⫹ 2y3 ⱕ 10 Dual constraints
y1 ⫹ 2y2 ⫹ 2y3 ⱕ 8
where y1 ⱖ 0, y2 ⱖ 0, and y3 ⱖ 0. Now apply the simplex method to the dual problem,
as follows.
332600_09_4.qxp 4/22/08 4:52 PM Page 573
Basic
y1 y2 y3 s1 s2 s3 b Variables
1 0 ⫺1 1 0 0 2 s1
1 1 2 0 1 0 10 s2
1 2 2 0 0 1 8 s3 ← Departing
⫺6 ⫺8 ⫺4 0 0 0 0
→
Entering
Basic
1 0 ⫺1 1 0 0 2 s1 ← Departing
1
2 0 1 0 1 ⫺ 12 6 s2
1 1
2 1 1 0 0 2 4 y2
⫺2 0 4 0 0 4 32
→
Entering
Basic
1 0 ⫺1 1 0 0 2 y1
3
0 0 2 ⫺ 12 1 ⫺ 12 5 s2
3
⫺ 12 1
0 1 2 0 2 3 y2
0 0 2 2 0 4 36
→
→
→
x1 x2 x3
From this final simplex tableau, you can see that the maximum value of z is 36. So, the
solution of the original minimization problem is
w ⫽ 36 Minimum value
and this occurs when

x1 ⫽ 2, x2 ⫽ 0, and x3 ⫽ 4.
Application
EXAMPLE 3 A Business Application: Minimum Cost

A small petroleum company owns two refineries. Refinery 1 costs $20,000 per day to
operate, and it can produce 400 barrels of high-grade oil, 300 barrels of medium-grade oil,
and 200 barrels of low-grade oil each day. Refinery 2 is newer and more modern. It costs
332600_09_4.qxp 4/22/08 4:52 PM Page 574
$25,000 per day to operate, and it can produce 300 barrels of high-grade oil, 400 barrels of
medium-grade oil, and 500 barrels of low-grade oil each day.
The company has orders totaling 25,000 barrels of high-grade oil, 27,000 barrels of
medium-grade oil, and 30,000 barrels of low-grade oil. How many days should it run each
refinery to minimize its costs and still refine enough oil to meet its orders?
SOLUTION To begin, let x1 and x2 represent the numbers of days the two refineries are operated. Then
the total cost is represented by
C ⫽ 20,000x1 ⫹ 25,000x2. Objective function
The constraints are

(High-grade) 400x1 ⫹ 300x2 ⱖ 25,000
(Medium-grade) 300x1 ⫹ 400x2 ⱖ 27,000 Constraints
(Low-grade) 200x1 ⫹ 500x2 ⱖ 30,000
where x1 ⱖ 0 and x2 ⱖ 0. The augmented matrix corresponding to this minimization
problem is
⯗
冤冥
400 300 25,000
300 400 ⯗ 27,000
200 500 ⯗ 30,000 .
... ... ⯗ ...
20,000 25,000 ⯗ 0
The matrix corresponding to the dual maximization problem is
⯗
冤冥
400 300 200 20,000
300 400 500 ⯗ 25,000
... ... ... ⯗ ... .
25,000 27,000 30,000 ⯗ 0
Now apply the simplex method to the dual problem as follows.
Basic
400 300 200 1 0 20,000 s1
300 400 500 0 1 25,000 s2 ← Departing
⫺25,000 ⫺27,000 ⫺30,000 0 0 0

→
Entering
332600_09_4.qxp 4/22/08 4:52 PM Page 575
Basic
280 140 0 1 ⫺ 25 10,000 s1 ← Departing

3 4 1
5 5 1 0 500 50 y3
⫺7000 ⫺3000 0 0 60 1,500,000
→
Entering
Basic
1 1 1 250
1 2 0 280 ⫺ 700 7 y1
1 3 1 200
0 2 1 ⫺ 1400 350 7 y3
0 500 0 25 50 1,750,000
→
→
x1 x2
From the third simplex tableau, you can see that the solution to the original minimization
problem is
C ⫽ $1,750,000 Minimum cost
and this occurs when x1 ⫽ 25 and x2 ⫽ 50. So, the two refineries should be operated for
the following numbers of days.
Refinery 1: 25 days
Refinery 2: 50 days
Note that by operating the two refineries for these numbers of days, the company will have
produced the amounts of oil shown below.
High-grade oil: 25共400兲 ⫹ 50共300兲 ⫽ 25,000 barrels
Medium-grade oil: 25共300兲 ⫹ 50共400兲 ⫽ 27,500 barrels
Low-grade oil: 25共200兲 ⫹ 50共500兲 ⫽ 30,000 barrels
So, the original production level has been met (with a surplus of 500 barrels of
medium-grade oil).
332600_09_4.qxp 4/22/08 4:52 PM Page 576

In Exercises 1–6, determine the dual of the minimization problem. 9. Objective function: 10. Objective function:
1. Objective function: 2. Objective function: w ⫽ x1 ⫹ 4x2 w ⫽ 2x1 ⫹ 6x2
w ⫽ 3x1 ⫹ 3x2 w ⫽ 2x1 ⫹ x2 Constraints: Constraints:
Constraints: Constraints: x1 ⫹ x2 ⱖ 3 ⫺2x1 ⫹ 3x2 ⱖ 0
2x1 ⫹ x2 ⱖ 4 5x1 ⫹ x2 ⱖ 9 ⫺x1 ⫹ 2x2 ⱖ 2 x1 ⫹ 3x2 ⱖ 9
x1 ⫹ 2x2 ⱖ 4 2x1 ⫹ 2x2 ⱖ 10 x1, x2 ⱖ 0 x1, x2 ⱖ 0
x1, x2 ⱖ 0 x1, x2 ⱖ 0 x2 x2
3. Objective function: 4. Objective function: (0, 3)
3 10
w ⫽ 4x1 ⫹ x2 ⫹ x3 w ⫽ 9x1 ⫹ 6x2
8
Constraints: Constraints: 2
6
3x1 ⫹ 2x2 ⫹ x3 ⱖ 23 x1 ⫹ 2x2 ⱖ 5 (43, 53( 4 (0, 3)
x1 ⫹ x3 ⱖ 10 2x1 ⫹ 2x2 ⱖ 8 2 (3, 2)
8x1 ⫹ x2 ⫹ 2x3 ⱖ 40 2x1 ⫹ x2 ⱖ 6 x1 x1
1 2 3
2 4 6 8 10
x1, x2, x3 ⱖ 0 x1, x2 ⱖ 0
5. Objective function: 6. Objective function: 11. Objective function: 12. Objective function:
w ⫽ 14x1 ⫹ 20x2 ⫹ 24x3 w ⫽ 9x1 ⫹ 4x2 ⫹ 10x3 w ⫽ 6x1 ⫹ 3x2 w ⫽ x1 ⫹ 6x2
x1 ⫹ x2 ⫹ 2x3 ⱖ 7 2x1 ⫹ x2 ⫹ 3x3 ⱖ 6 4x1 ⫹ x2 ⱖ 4 2x1 ⫹ 3x2 ⱖ 15
x1 ⫹ 2x2 ⫹ x3 ⱖ 4 6x1 ⫹ x2 ⫹ x3 ⱖ 9 x2 ⱖ 2 ⫺x1 ⫹ 2x2 ⱖ 3
x1, x2, x3 ⱖ 0 x1, x2, x3 ⱖ 0 x1, x2 ⱖ 0 x1, x2 ⱖ 0
In Exercises 7–12, (a) solve the minimization problem by the graph- x2 x2
ical method, (b) formulate the dual problem, and (c) solve the dual
8
problem by the graphical method.
(0, 4)
4 6
7. Objective function: 8. Objective function: (0, 5)
3
w ⫽ 2x1 ⫹ 2x2 w ⫽ 14x1 ⫹ 20x2 4
(3, 3)
1 ( 2(
1
2
, 2
x1 ⫹ 2x2 ⱖ 3 x1 ⫹ 2x2 ⱖ 4
x1 x1
3x1 ⫹ 2x2 ⱖ 5 7x1 ⫹ 6x2 ⱖ 20 2 3 4 5 2 4 6
x1, x2 ⱖ 0 x1, x2 ⱖ 0
In Exercises 13–24, solve the minimization problem by solving the
x2 x2 dual maximization problem with the simplex method.

(0, 52 ( (0, 103( w ⫽ x2 w ⫽ 3x1 ⫹ 8x2
3
2
(1, 1)
1 (2, 1) x1 ⫹ 5x2 ⱖ 10 2x1 ⫹ 7x2 ⱖ 9
1
(3, 0) (4, 0) ⫺6x1 ⫹ 5x2 ⱖ 3 x1 ⫹ 2x2 ⱖ 4
x1 x1
1 3 1 2 3 4 x1, x2 ⱖ 0 x1, x2 ⱖ 0
332600_09_4.qxp 4/22/08 4:52 PM Page 577
15. Objective function: 16. Objective function: 25. The first drink costs $2 per liter and the second costs $3 per liter.
w ⫽ 2x1 ⫹ x2 w ⫽ 2x1 ⫹ 2x2 26. The first drink costs $4 per liter and the second costs $2 per liter.
Constraints: Constraints: 27. The first drink costs $1 per liter and the second costs $3 per liter.
5x1 ⫹ x2 ⱖ 9 3x1 ⫹ x2 ⱖ 6 28. The first drink costs $1 per liter and the second costs $2 per liter.
2x1 ⫹ 2x2 ⱖ 10 ⫺4x1 ⫹ 2x2 ⱖ 2 In Exercises 29–32, an athlete uses two dietary supplement drinks
x1, x2 ⱖ 0 x1, x2 ⱖ 0 that provide the nutritional elements shown in the table.
Drink Protein Carbohydrates Vitamin D
w ⫽ 4x1 ⫹ x2 w ⫽ x1 ⫹ x2
Constraints: Constraints: I 4 2 1
x1 ⫹ x2 ⱖ 8 x1 ⫹ 2x2 ⱖ 40
II 1 5 1
3x1 ⫹ 5x2 ⱖ 30 2x1 ⫹ 3x2 ⱖ 72
x1, x2 ⱖ 0 x1, x2 ⱖ 0 Find the combination of drinks of minimum cost that will meet the
19. Objective function: 20. Objective function: minimum requirements of 4 units of protein, 10 units of carbohy-
w ⫽ x1 ⫹ 2x2 w ⫽ 3x1 ⫹ 4x2 drates, and 3 units of vitamin D.
Constraints: Constraints: 29. Drink I costs $5 per liter and drink II costs $8 per liter.
3x1 ⫹ 5x2 ⱖ 15 x1 ⫹ 2x2 ⱖ 6 30. Drink I costs $7 per liter and drink II costs $4 per liter.
5x1 ⫹ 2x2 ⱖ 10 2x1 ⫹ x2 ⱖ 10 31. Drink I costs $1 per liter and drink II costs $5 per liter.
x1, x2 ⱖ 0 x1, x2 ⱖ 0 32. Drink I costs $8 per liter and drink II costs $1 per liter.
21. Objective function: 22. Objective function: 33. An electronics manufacturing company has three production
w ⫽ 8x1 ⫹ 4x2 ⫹ 6x3 w ⫽ 8x1 ⫹ 16x2 ⫹ 18x3 plants, each of which produces three different models of a
Constraints: Constraints: particular MP3 player. The daily capacities (in thousands of
units) of the three plants are shown in the table.
3x1 ⫹ 2x2 ⫹ x3 ⱖ 6 2x1 ⫹ 2x2 ⫺ 2x3 ⱖ 4
4x1 ⫹ x2 ⫹ 3x3 ⱖ 7 ⫺4x1 ⫹ 3x2 ⫺ x3 ⱖ 1 Basic Gold Platinum
2x1 ⫹ x2 ⫹ 4x3 ⱖ 8 x1 ⫺ x2 ⫹ 3x3 ⱖ 8 Model Model Model
x1, x2, x3 ⱖ 0 x1, x2, x3 ⱖ 0
Plant 1 8 4 8
w ⫽ 6x1 ⫹ 2x2 ⫹ 3x3 w ⫽ 42x1 ⫹ 5x2 ⫹ 17x3 Plant 2 6 6 3
Constraints: Constraints: Plant 3 12 4 8
3x1 ⫹ 2x2 ⫹ x3 ⱖ 28 3x1 ⫺ x2 ⫹ 7x3 ⱖ 5
6x1 ⫹ x3 ⱖ 24 ⫺3x1 ⫺ x2 ⫹ 3x3 ⱖ 8 The total demands are 300,000 units of the Basic model,
3x1 ⫹ x2 ⫹ 2x3 ⱖ 40 6x1 ⫹ x2 ⫹ x3 ⱖ 16 172,000 units of the Gold model, and 249,500 units of the
Platinum model. The daily operating costs are $55,000 for
x1, x2, x3 ⱖ 0 x1, x2, x3 ⱖ 0
plant 1, $60,000 for plant 2, and $60,000 for plant 3. How many
In Exercises 25–28, two sports drinks are used to supply protein and days should each plant be operated in order to fill the total
carbohydrates. The first drink provides 1 unit of protein and 3 units demand while keeping the operating cost at a minimum?
of carbohydrates in each liter. The second drink supplies 2 units of 34. The company in Exercise 33 has lowered the daily operating
protein and 2 units of carbohydrates in each liter. An athlete cost for plant 3 to $50,000. How many days should each plant
requires 3 units of protein and 5 units of carbohydrates. Find the be operated in order to fill the total demand while keeping the
amount of each drink the athlete should consume to minimize the operating cost at a minimum?
cost and still meet the minimum dietary requirements.
332600_09_4.qxp 4/22/08 4:52 PM Page 578
35. A small petroleum company owns two refineries. Refinery 1 38. Use a computer software program to minimize the objective
costs $25,000 per day to operate, and it can produce 300 barrels function
of high-grade oil, 200 barrels of medium-grade oil, and 150 w ⫽ 1.5x1 ⫹ x2 ⫹ 0.5x3 ⫹ 2x4
barrels of low-grade oil each day. Refinery 2 is newer and more
subject to the same set of constraints for Exercise 37.
modern. It costs $30,000 per day to operate, and it can produce
300 barrels of high-grade oil, 250 barrels of medium-grade oil, True or False? In Exercises 39 and 40, determine whether each
and 400 barrels of low-grade oil each day. The company has statement is true or false. If a statement is true, give a reason or cite
orders totaling 35,000 barrels of high-grade oil, 30,000 barrels an appropriate statement from the text. If a statement is false,
of medium-grade oil, and 40,000 barrels of low-grade oil. How provide an example that shows the statement is not true in all cases
many days should the company run each refinery to minimize its or cite an appropriate statement from the text.
costs and still meet its orders?
39. The basic procedure for solving a minimization problem is to
36. A steel company has two mills. Mill 1 costs $70,000 per day to convert it to a maximization problem and then apply the simplex
operate, and it can produce 400 tons of high-grade steel, 500 method.
tons of medium-grade steel, and 450 tons of low-grade steel
40. The corresponding maximization problem is called the dual of
each day. Mill 2 costs $60,000 per day to operate, and it can
the original minimization problem.
produce 350 tons of high-grade steel, 600 tons of medium-grade
steel, and 400 tons of low-grade steel each day. The company
has orders totaling 100,000 tons of high-grade steel, 150,000
tons of medium-grade steel, and 124,500 tons of low-grade
steel. How many days should the company run each mill to
minimize its costs and still fill the orders?
37. Use a computer software program to minimize the objective
function
w ⫽ x1 ⫹ 0.5x2 ⫹ 2.5x3 ⫹ 3x4
1.5x1 ⫹ x2 ⫹ 2x4 ⱖ 35
2x2 ⫹ 6x3 ⫹ 4x4 ⱖ 120
x1 ⫹ x2 ⫹ x3 ⫹ x4 ⱖ 50
0.5x1 ⫹ 2.5x3 ⫹ 1.5x4 ⱖ 75
where x1, x2, x3, x4 ⱖ 0.
332600_09_5.qxp 4/22/08 4:53 PM Page 579
Section 9.5 The Simplex Method: Mixed Constraints 579
9.5 The Simplex Method: Mixed Constraints

In Sections 9.3 and 9.4, you looked at linear programming problems that occurred in
standard form. The constraints for the maximization problems all involved ⱕ inequalities,
and the constraints for the minimization problems all involved ⱖ inequalities.
Linear programming problems for which the constraints involve both types of
inequalities are called mixed-constraint problems. For instance, consider the following
linear programming problem.
Mixed-Constraint Problem: Find the maximum value of
z ⫽ x1 ⫹ x2 ⫹ 2x3 Objective function

2x1 ⫹ x2 ⫹ x3 ⱕ 50
2x1 ⫹ x2 ⱖ 36 Constraints
x1 ⫹ x3 ⱖ 10
where x1 ⱖ 0, x2 ⱖ 0, and x3 ⱖ 0. Because this is a maximization problem, you would
expect each of the inequalities in the set of constraints to involve ⱕ. Moreover, because the
first inequality does involve ⱕ, you can add a slack variable to form the equation
2x1 ⫹ x2 ⫹ x3 ⫹ s1 ⫽ 50.
For the other two inequalities, a new type of variable, called a surplus variable, is
introduced as follows.
2x1 ⫹ x2 ⫺ s2 ⫽ 36
x1 ⫹ x3 ⫺ s3 ⫽ 10
Notice that surplus variables are subtracted from (not added to) their inequalities. The
variables s2 and s3 are called surplus variables because they represent the amounts by
which the left sides of the inequalities exceed the right sides. Surplus variables must be
nonnegative.
Now, to solve the linear programming problem, form an initial simplex tableau, as
follows.
Basic
2 1 1 1 0 0 50 s1
2 1 0 0 ⫺1 0 36 s2
1 0 1 0 0 ⫺1 10 s3 ← Departing
⫺1 ⫺1 ⫺2 0 0 0 0
→
Entering
332600_09_5.qxp 4/22/08 4:53 PM Page 580
You will soon discover that solving mixed-constraint problems can be difficult. One
reason for this is that there is no convenient feasible solution to begin the simplex method.
Note that the solution represented by the initial tableau above,
共x1, x2, x3, s1, s2, s3兲 ⫽ 共0, 0, 0, 50, ⫺36, ⫺10兲,
is not a feasible solution because the values of the two surplus variables are negative. In
fact, the values x1 ⫽ x2 ⫽ x3 ⫽ 0 do not even satisfy the constraint equations. In order to
eliminate the surplus variables from the current solution, “trial and error” is used. That is,
in an effort to find a feasible solution, arbitrarily choose new entering variables. For
instance, in this tableau, it seems reasonable to select x3 as the entering variable. After
pivoting, the new simplex tableau is as follows.
Basic
1 1 0 1 0 1 40 s1
2 1 0 0 ⫺1 0 36 s2 ← Departing
1 0 1 0 0 ⫺1 10 x3
1 ⫺1 0 0 0 ⫺2 20
→
Entering
The current solution 共x1, x2, x3, s1, s2, s3兲 ⫽ 共0, 0, 10, 40, ⫺36, 0兲 is still not feasible, so
choose x2 as the entering variable and pivot to obtain the following simplex tableau.
Basic
⫺1 0 0 1 1 1 4 s1 ← Departing
2 1 0 0 ⫺1 0 36 x2
1 0 1 0 0 ⫺1 10 x3
3 0 0 0 ⫺1 ⫺2 56
→
Entering
At this point, the feasible solution shown below is finally obtained.

共x1, x2, x3, s1, s2, s3兲 ⫽ 共0, 36, 10, 4, 0, 0兲
From here on, you can apply the simplex method as usual. Note that the entering variable
here is s3 because its column has the most negative entry in the bottom row. After pivoting
one more time, you obtain the final simplex tableau shown on the next page.
332600_09_5.qxp 4/22/08 4:53 PM Page 581
Basic
⫺1 0 0 1 1 1 4 s3
2 1 0 0 ⫺1 0 36 x2
0 0 1 1 1 0 14 x3
1 0 0 2 1 0 64
Note that this tableau is final because it represents a feasible solution and there are no
negative entries in the bottom row. So, you can conclude that the maximum value of the
objective function is
z ⫽ 64 Maximum value

x1 ⫽ 0, x2 ⫽ 36, and x3 ⫽ 14.
EXAMPLE 1 A Maximization Problem with Mixed Constraints

z ⫽ 3x1 ⫹ 2x2 ⫹ 4x3 Objective function

3x1 ⫹ 2x2 ⫹ 5x3 ⱕ 18
4x1 ⫹ 2x2 ⫹ 3x3 ⱕ 16 Constraints
2x1 ⫹ x2 ⫹ x3 ⱖ 4
SOLUTION To begin, add a slack variable to each of the first two inequalities and subtract a surplus
variable from the third inequality to produce the initial simplex tableau shown below.
Basic
3 2 5 1 0 0 18 s1
4 2 3 0 1 0 16 s2
2 1 1 0 0 ⫺1 4 s3 ← Departing
⫺3 ⫺2 ⫺4 0 0 0 0
→
Entering
332600_09_5.qxp 4/22/08 4:53 PM Page 582
As it stands, this tableau does not represent a feasible solution (because the value of s3 is
negative). So, s3 should be the departing variable. There are no real guidelines as to which
variable should enter the solution, but by trial and error, you will discover that using x2 as
the entering variable produces the tableau below (which does represent a feasible solution).
Basic
⫺1 0 3 1 0 2 10 s1 ← Departing
0 0 1 0 1 2 8 s2
2 1 1 0 0 ⫺1 4 x2
1 0 ⫺2 0 0 ⫺2 8
Now, because this simplex tableau does represent a feasible solution, you can proceed as
usual, choosing the most negative entry in the bottom row to be the entering variable. (In
this case, there is a tie, so arbitrarily choose x3 to be the entering variable.)
Basic
⫺1 0 3 1 0 2 10 s1 ← Departing
0 0 1 0 1 2 8 s2
2 1 1 0 0 ⫺1 4 x2
1 0 ⫺2 0 0 ⫺2 8
→
Entering
Basic
⫺ 13 1 2 10
0 1 3 0 3 3 x3
1
⫺ 13 4 14
3 0 0 1 3 3 s2 ← Departing
7
⫺ 13 ⫺ 53 2
3 1 0 0 3 x2
1 2
3 0 0 3 0 ⫺ 23 44
3
→
Entering
Basic
1 1 1
⫺2 0 1 2 ⫺2 0 1 x3
1
4 0 0 ⫺ 14 3
4 1 7
2 s3
11
⫺ 34 5 13
4 1 0 4 0 2 x2
1 1 1
2 0 0 2 2 0 17
332600_09_5.qxp 4/22/08 4:53 PM Page 583
So, the maximum value of the objective function is

z ⫽ 17
x1 ⫽ 0, x2 ⫽ 13
2 , and x3 ⫽ 1.
Mixed Constraints and Minimization

In Section 9.4, the solution of minimization problems in standard form was discussed.
Minimization problems that are not in standard form are more difficult to solve. One
technique that can be used is to change a mixed-constraint minimization problem to a
mixed-constraint maximization problem by multiplying each coefficient in the objective
function by ⫺1. This technique is demonstrated in the next example.
EXAMPLE 2 A Minimization Problem with Mixed Constraints

w ⫽ 4x1 ⫹ 2x2 ⫹ x3 Objective function

2x1 ⫹ 3x2 ⫹ 4x3 ⱕ 14
3x1 ⫹ x2 ⫹ 5x3 ⱖ 4 Constraints
x1 ⫹ 4x2 ⫹ 3x3 ⱖ 6
SOLUTION First, rewrite the objective function by multiplying each of its coefficients by ⫺1, as
follows.
z ⫽ ⫺4x1 ⫺ 2x2 ⫺ x3 Revised objective function
Maximizing this revised objective function is equivalent to minimizing the original objec-
tive function. Next, add a slack variable to the first inequality and subtract surplus variables
from the second and third inequalities to produce the initial simplex tableau shown below.
Basic
2 3 4 1 0 0 14 s1
3 1 5 0 ⫺1 0 4 s2 ← Departing
1 4 3 0 0 ⫺1 6 s3
4 2 1 0 0 0 0
→
Entering
332600_09_5.qxp 4/22/08 4:53 PM Page 584
Note that the bottom row contains the negatives of the coefficients of the revised objective
function. Another way of looking at this is that for minimization problems (in nonstandard
form), the bottom row of the initial simplex tableau consists of the coefficients of the
original objective function.
As with maximization problems with mixed constraints, this initial simplex tableau does
not represent a feasible solution. By trial and error, you will discover that you can choose
x2 as the entering variable and s2 as the departing variable. After pivoting, you obtain the
tableau shown below.
Basic
⫺7 0 ⫺11 1 3 0 2 s1
3 1 5 0 ⫺1 0 4 x2
⫺11 0 ⫺17 0 4 ⫺1 ⫺10 s3
⫺2 0 ⫺9 0 2 0 ⫺8
From this tableau, you can see that the choice of x2 as the entering variable was
appropriate. All you need to do to transform the tableau into one that represents a feasible
solution is to multiply the third row by ⫺1, as follows.
Basic
⫺7 0 ⫺11 1 3 0 2 s1
3 1 5 0 ⫺1 0 4 x2
11 0 17 0 ⫺4 1 10 s3 ← Departing
⫺2 0 ⫺9 0 2 0 ⫺8
→
Entering
Now that you have obtained a simplex tableau that represents a feasible solution, continue
with the standard pivoting operations, as follows.
Basic
2 7 11 144
17 0 0 1 17 17 17 s1
4 3 5 18
⫺ 17 1 0 0 17 ⫺ 17 17 x2 ← Departing
11 4 1 10
17 0 1 0 ⫺ 17 17 17 x3
65 2 9
17 0 0 0 ⫺ 17 17 ⫺ 46
17
→
Entering
332600_09_5.qxp 4/22/08 4:53 PM Page 585
Basic
2
⫺ 73 4
3 0 1 0 3 6 s1
⫺ 43 17
3 0 0 1 ⫺ 53 6 s2
1 4
3 3 1 0 0 ⫺ 13 2 x3
11 2 1
3 3 0 0 0 3 ⫺2
Finally, you can conclude that the maximum value of the revised objective function is
z ⫽ ⫺2, and so the minimum value of the original objective function is
w⫽2
(the negative of the entry in the lower-right corner). This occurs when
x1 ⫽ 0, x2 ⫽ 0, and x3 ⫽ 2.
Application
EXAMPLE 3 A Business Application: Minimum Shipment Costs

An automobile company has two factories. One factory has 400 cars (of a certain model)
in stock and the other factory has 300 cars (of the same model) in stock. Two customers
order this car model. The first customer needs 200 cars, and the second customer needs
300 cars. The costs of shipping cars from the two factories to the customers are shown in
Table 9.3.
TABLE 9.3
Customer 1 Customer 2
Factory 1 $30 $25

Factory 2 $36 $30
How should the company ship the cars in order to minimize the shipping costs?
SOLUTION To begin, let x1 and x2 represent the numbers of cars shipped from factory 1 to the first and
second customers, respectively. (See Figure 9.20.) The total cost of shipping is
C ⫽ 30x1 ⫹ 25x2 ⫹ 36共200 ⫺ x1兲 ⫹ 30共300 ⫺ x2兲 ⫽ 16,200 ⫺ 6x1 ⫺ 5x2.
332600_09_5.qxp 4/22/08 4:53 PM Page 586
Factory 1
The constraints for this minimization problem are as follows.
x1 ⫹ x2 ⱕ 400
$30 $25
x1 x2 共200 ⫺ x1兲 ⫹ 共300 ⫺ x2兲 ⱕ 300 x1 ⫹ x2 ⱖ 200
x1 ⱕ 200
Customer 1 Customer 2 x2 ⱕ 300
300 − x2 The corresponding maximization problem is to maximize z ⫽ 6x1 ⫹ 5x2 ⫺ 16,200. So, the
200 − x1 initial simplex tableau is as follows.
$36 $30
Basic
Factory 2 x1 x2 s1 s2 s3 s4 b Variables
Figure 9.20 1 1 1 0 0 0 400 s1

1 1 0 ⫺1 0 0 200 s2 ← Departing
1 0 0 0 1 0 200 s3
0 1 0 0 0 1 300 s4
⫺6 ⫺5 0 0 0 0 ⫺16,200
→
Entering
Note that the current z-value is ⫺16,200 because the initial solution is
共x1, x2, s1, s2, s3, s4兲 ⫽ 共0, 0, 400, ⫺200, 200, 300兲.
Now, to this initial tableau, apply the simplex method as follows.
Basic
x1 x2 s1 s2 s3 s4 b Variables
0 0 1 1 0 0 200 s1
1 1 0 ⫺1 0 0 200 x1
0 ⫺1 0 1 1 0 0 s3 ← Departing
0 1 0 0 0 1 300 s4
0 1 0 ⫺6 0 0 ⫺15,000
→
Entering
332600_09_5.qxp 4/22/08 4:53 PM Page 587
Basic
0 1 1 0 ⫺1 0 200 s1 ← Departing
1 0 0 0 1 0 200 x1
0 ⫺1 0 1 1 0 0 s2
0 1 0 0 0 1 300 s4
0 ⫺5 0 0 6 0 ⫺15,000
→
Entering
Basic
0 1 1 0 ⫺1 0 200 x2
1 0 0 0 1 0 200 x1
0 0 1 1 0 0 200 s2
0 0 ⫺1 0 1 1 100 s4
0 0 5 0 1 0 ⫺14,000
From this tableau, you can see that the minimum shipping cost is $14,000. Because
x1 ⫽ 200 and x2 ⫽ 200, you can conclude that the numbers of cars that should be shipped
from the two factories are as shown in Table 9.4.
TABLE 9.4
Customer 1 Customer 2
Factory 1 200 cars 200 cars

Factory 2 0 100 cars
332600_09_5.qxp 4/22/08 4:53 PM Page 588

In Exercises 1–6, add the appropriate slack and surplus variables to 9. (Minimize) 10. (Minimize)
the system and form the initial simplex tableau. Objective function: Objective function:
1. (Maximize) 2. (Maximize) w ⫽ x1 ⫹ 2x2 w ⫽ 3x1 ⫹ 2x2
Objective function: Objective function: Constraints: Constraints:
w ⫽ 10x1 ⫹ 4x2 w ⫽ 3x1 ⫹ x2 ⫹ x3 2x1 ⫹ 3x2 ⱕ 25 x1 ⫹ x2 ⱖ 20
Constraints: Constraints: x1 ⫹ 2x2 ⱖ 16 3x1 ⫹ 4x2 ⱕ 70
2x1 ⫹ x2 ⱖ 4 x1 ⫹ 2x2 ⫹ x3 ⱕ 10 x1, x2 ⱖ 0 x1, x2 ⱖ 0
x1 ⫹ x2 ⱕ 8 x2 ⫹ 5x3 ⱖ 6 Entering x2, departing s2 Entering x1, departing s1
x1, x2 ⱖ 0 4x1 ⫺ x2 ⫹ x3 ⱖ 16 11. (Maximize) 12. (Maximize)
x1, x2, x3 ⱖ 0 Objective function: Objective function:
3. (Minimize) 4. (Minimize) w ⫽ x1 ⫹ x2 w ⫽ x1 ⫹ 2x2 ⫹ 2x3
Objective function: Objective function: Constraints: Constraints:
w ⫽ x1 ⫹ x2 w ⫽ 2x1 ⫹ 3x2 ⫺4x1 ⫹ 3x2 ⫹ x3 ⱕ 40 x1 ⫹ x2 ⱖ 50
Constraints: Constraints: ⫺2x1 ⫹ x2 ⫹ x3 ⱖ 10 2x1 ⫹ x2 ⫹ x3 ⱕ 70
2x1 ⫹ x2 ⱕ 4 3x1 ⫹ x2 ⱖ 4 x2 ⫹ x3 ⱕ 20 x2 ⫹ 3x3 ⱖ 40
x1 ⫹ 3x2 ⱖ 2 4x1 ⫹ 2x2 ⱕ 3 x1, x2, x3 ⱖ 0 x1, x2, x3 ⱖ 0
x1, x2 ⱖ 0 x1, x2 ⱖ 0 Entering x2, departing s2 Entering x2, departing s1
5. (Maximize) 6. (Maximize) In Exercises 13–20, use the simplex method to solve the problem.
w ⫽ x1 ⫹ x3 w ⫽ 4x1 ⫹ x2 ⫹ x3
w ⫽ 2x1 ⫹ 5x2 w ⫽ ⫺x1 ⫹ 3x2
4x1 ⫹ x2 ⱖ 10 2x1 ⫹ x2 ⫹ 4x3 ⱕ 60
x1 ⫹ x2 ⫹ 3x3 ⱕ 30 x2 ⫹ x3 ⱖ 40
x1 ⫹ 2x2 ⱖ 4 2x1 ⫹ x2 ⱕ 4
2x1 ⫹ x2 ⫹ 4x3 ⱖ 16 x1, x2, x3 ⱖ 0
x1 ⫹ x2 ⱕ 8 x1 ⫹ 5x2 ⱖ 5
x1, x2, x3 ⱖ 0
x1, x2 ⱖ 0 x1, x2 ⱖ 0
In Exercises 7–12, use the specified entering and departing
variables to solve the mixed-constraint problem.
w ⫽ 2x1 ⫹ x2 ⫹ 3x3 w ⫽ 3x1 ⫹ 5x2 ⫹ 2x3
w ⫽ ⫺x1 ⫹ 2x2 w ⫽ 2x1 ⫹ x2
x1 ⫹ 4x2 ⫹ 2x3 ⱕ 85 9x1 ⫹ 4x2 ⫹ x3 ⱕ 70
x2 ⫺ 5x3 ⱖ 20 5x1 ⫹ 2x2 ⫹ x3 ⱕ 40
x1 ⫹ x2 ⱖ 3 x1 ⫹ x2 ⱖ 4
3x1 ⫹ 2x2 ⫹ 11x3 ⱖ 49 4x1 ⫹ x2 ⱖ 16
x1 ⫹ x2 ⱕ 6 x1 ⫹ x2 ⱕ 8
x1, x2, x3 ⱖ 0 x1, x2, x3 ⱖ 0
x1, x2 ⱖ 0 x1, x2 ⱖ 0
Entering x2, departing s1 Entering x1, departing s1
332600_09_5.qxp 4/22/08 4:53 PM Page 589
17. (Minimize) 18. (Minimize) 29. 30.

C1 C2 C1 C2
w ⫽ x1 ⫹ x2 w ⫽ 2x1 ⫹ 3x2 S1 0.60 1.20 S1 0.80 1.00
Constraints: Constraints: S2 1.00 1.80 S2 1.00 1.20
x1 ⫹ 2x2 ⱖ 25 3x1 ⫹ 2x2 ⱕ 22
2x1 ⫹ 5x2 ⱕ 60 x1 ⫹ x2 ⱖ 10 31. 32.
C1 C2 C1 C2
x1, x2 ⱖ 0 x1, x2 ⱖ 0
19. (Minimize) 20. (Minimize) S1 1.20 1.00 S1 0.80 1.00
Objective function: Objective function: S2 1.00 1.20 S2 1.00 0.80
w ⫽ ⫺2x1 ⫹ 4x2 ⫺ x3 w ⫽ x1 ⫹ x2 ⫹ x3
Constraints: Constraints: 33. An automobile company has two factories. One factory has 400
3x1 ⫺ 6x2 ⫹ 4x3 ⱕ 30 x1 ⫹ 2x2 ⫹ x3 ⱖ 30 cars (of a certain model) in stock and the other factory has 300
cars (of the same model) in stock. Two customers order this car
2x1 ⫺ 8x2 ⫹ 10x3 ⱖ 18 6x2 ⫹ x3 ⱕ 54 model. The first customer needs 200 cars, and the second
x1, x2, x3 ⱖ 0 x1 ⫹ x2 ⫹ 3x3 ⱖ 20 customer needs 300 cars. The costs of shipping cars from the
x1, x2, x3 ⱖ 0 two factories to the two customers are shown in the table.
In Exercises 21–24, maximize the objective function subject to the Customer 1 Customer 2
constraints listed below.
Factory 1 $36 $30
x1 ⫹ x2 ⱕ 5
Factory 2 $30 $25
⫺x1 ⫹ x2 ⱕ 3
x2 ⱖ 1
How should the company ship the cars in order to minimize the
x1, x2 ⱖ 0 shipping costs?
21. w ⫽ 2x1 ⫹ x2 22. w ⫽ x1 ⫹ 2x2 34. Suppose the shipping costs for the two factories in Exercise 33
are as shown in the table below.
23. w ⫽ x2 24. w ⫽ ⫺x1 ⫺ x2
In Exercises 25–28, maximize the objective function subject to the Customer 1 Customer 2
constraints listed below.
Factory 1 $25 $30
3x1 ⫹ 2x2 ⱖ 6 Factory 2 $35 $30
x1 ⫺ x2 ⱕ 2
⫺x1 ⫹ 2x2 ⱕ 6 How should the company ship the cars in order to minimize the
x1 ⱕ4 shipping costs?
x1, x2 ⱖ 0 35. A company has budgeted a maximum of $600,000 for advertis-
ing a certain product nationally. Each minute of television time
25. w ⫽ x1 ⫹ x2 26. w ⫽ x1 ⫺ 2x2 costs $60,000 and each one-page newspaper ad costs $15,000.
27. w ⫽ ⫺4x1 ⫹ x2 28. w ⫽ 4x1 ⫺ x2 Each television ad is expected to be viewed by 15 million view-
ers, and each newspaper ad is expected to be seen by 3 million
In Exercises 29–32, a tire company has two suppliers, S1 and S2. S1 readers. The company’s market research department advises the
has 900 tires on hand and S2 has 800 tires on hand. Customer C1 company to use at least 6 television ads and at least 4 newspaper
needs 500 tires and customer C2 needs 600 tires. Minimize the cost ads. How should the advertising budget be allocated to maxi-
of filling the orders subject to the data in the table (shipping costs mize the total audience?
per tire).
332600_09_5.qxp 4/22/08 4:53 PM Page 590
36. Rework Exercise 35 assuming that each one-page newspaper ad True or False? In Exercises 39–42, determine whether each state-
costs $30,000. ment is true or false. If a statement is true, give a reason or cite an
appropriate statement from the text. If a statement is false, provide
In Exercises 37 and 38, use the given information. A computer
an example that shows the statement is not true in all cases or cite
company has two assembly plants, plant A and plant B, and two
an appropriate statement from the text.
distribution outlets, outlet I and outlet II. Plant A can assemble 5000
computers in a year and plant B can assemble 4000 computers in a 39. Linear programming problems for which constraints contain
year. Outlet I must have 3000 computers per year and outlet II must both greater than or equal to and less than or equal to signs are
have 5000 computers per year. The costs of transportation from called mixed-constraint problems.
each plant to each outlet are shown in the table. Find the shipping 40. Surplus variables are subtracted from (not added to) inequalities
schedule that will produce the minimum cost. What is the minimum because they represent an amount for which the left side of the
cost? inequality exceeds the right side.
37. 41. One technique that can be used to change a mixed-constraint
Outlet I Outlet II minimization problem to a mixed-constraint maximization
problem is to multiply each coefficient of the objective function
Plant A $4 $5 by ⫺1.
Plant B $5 $6 42. Surplus variables, like slack variables, are always positive
because they represent the amount by which the left side of the
38. inequality is less than the right side.
Outlet I Outlet II
Plant A $4 $5
Plant B $6 $4
CHAPTER 9 Review Exercises

In Exercises 1–6, sketch a graph of the solution of the system of In Exercises 7 and 8, determine a system of inequalities that models
inequalities. the description, and sketch a graph of the solution of the system.
1. x ⫹ 2y ⱕ 160 2. 2x ⫹ 3y ⱕ 24 7. A Pennsylvania fruit grower has 1500 bushels of apples that are
3x ⫹ y ⱕ 180 2x ⫹ y ⱕ 16 to be divided between markets in Harrisburg and Philadelphia.
x ⱖ 0 x ⱖ 0 These two markets need at least 400 bushels and 600 bushels,
respectively.
y ⱖ 0 y ⱖ 0
8. A warehouse operator has 24,000 square meters of floor space
3. 3x ⫹ 2y ⱖ 24 4. 2x ⫹ y ⱖ 16 in which to store two products. Each unit of product I requires
x ⫹ 2y ⱖ 12 x ⫹ 3y ⱖ 18 20 square meters of floor space and costs $12 per day to store.
2 ⱕ x ⱕ 15 0 ⱕ x ⱕ 25 Each unit of product II requires 30 square meters of floor space
y ⱕ 15 0 ⱕ y ⱕ 15 and costs $8 per day to store. The total storage cost per day
cannot exceed $12,400.
5. 2x ⫺ 3y ⱖ 0 6. x ⫺ y ⱕ 10
2x ⫺ y ⱕ 8 x ⱖ 0
y ⱖ 0 y ⱖ 0
332600_09_5.qxp 4/22/08 4:53 PM Page 591
Chapter 9 Review E xercises 591
In Exercises 9–20, find the minimum and/or maximum values of the y y

objective function by the graphical method.
60 (0, 45)
800 (0, 800)
9. Maximize: z ⫽ 3x ⫹ 4y 10. Minimize: z ⫽ 10x ⫹ 7y (30, 45)
40 (0, 400)
y y 400 (900, 0)
20 (60, 20)
10
(0, 10)
100 (0, 100) (0, 0) (60, 0) x
8
(5, 8) x
75 400 (450, 0)
6 20 40 60
50 (25, 50)
4
2 25 (75, 0) 15. Objective function: 16. Objective function:
(0, 0) (7, 0)
x x
2 4 6 8 10 25 50 75 100 z ⫽ 5x ⫹ 0.5y z ⫽ 2x ⫹ y
−2 −25
11. Objective function: 12. Objective function: x ⱖ 0 x ⱖ 0
z ⫽ 4x ⫹ 3y z ⫽ 2x ⫹ 8y y ⱖ 0 2x ⫹ 3y ⱖ 6
Constraints: Constraints: x ⫹ 3y ⱕ 15 3x ⫺ y ⱕ 9
x ⱖ 0 x ⱖ 0 4x ⫹ y ⱕ 16 x ⫹ 4y ⱕ 16
y ⱖ 0 y ⱖ 0 y y
x⫹y ⱕ 5 2x ⫺ y ⱕ 4 5 (0, 5) 5
(3, 4) (0, 4)
y y 4 4
(4, 3)
3 3
6 4 (0, 4) (0, 2)
2 2
5 (0, 5) (4, 0)
3 1 1
4 (0, 0) (3, 0)
3 x x
2
1 2 3 4 5 1 2 3 4 5
2
1 (0, 0) (5, 0) (0, 0) (2, 0)
x x 17. Objective function: 18. Objective function:
1 2 3 4 5 6 −1 1 2 3
z ⫽ x ⫹ 3y z ⫽ 4x ⫺ y
z ⫽ 25x ⫹ 30y z ⫽ 15x ⫹ 20y x ⱖ 0 x ⱖ 0
Constraints: Constraints: y ⱖ 0 y ⱖ 0
x ⱖ 0 x ⱖ 0 x ⱕ 5 x ⱕ 6
y ⱖ 0 y ⱖ 0 x⫹y ⱖ 3 x⫹y ⱖ 2
x ⱕ 60 8x ⫹ 9y ⱕ 7200 x⫺y ⱕ 3 x ⱖ y
y ⱕ 45 8x ⫹ 9y ⱖ 3600 ⫺x ⫹ y ⱕ 3 3x ⫺ y ⱕ 12
5x ⫹ 6y ⱕ 420
332600_09_5.qxp 4/22/08 4:53 PM Page 592
19. Objective function: 20. Objective function: In Exercises 29 and 30, determine the dual of the minimization
z ⫽ 3x ⫺ y z ⫽ x ⫺ 2y problem.
x ⱖ 0 x ⱖ 0 w ⫽ 7x1 ⫹ 3x2 ⫹ x3 w ⫽ 2x1 ⫹ 3x2 ⫹ 4x3
y ⱖ 0 y ⱖ 0 Constraints: Constraints:
x ⱕ 3y x ⫹ y ⱕ 12 x1 ⫹ x2 ⫹ 2x3 ⱖ 30 x1 ⫹ 5x2 ⫹ 3x3 ⱖ 90
⫺x ⫹ 2y ⱕ 12 5x ⫹ y ⱕ 36 3x1 ⫹ 6x2 ⫹ 4x3 ⱖ 75 2x1 ⫹ x2 ⫹ 3x3 ⱖ 60
4x ⫹ 3y ⱕ 40 5x ⫺ 2y ⱖ 4 x1, x2, x3 ⱖ 0 3x1 ⫹ 2x2 ⫹ x3 ⱖ 56
x ⫹ 2y ⱕ 15 2x ⫹ 5y ⱖ 19 x1, x2, x3 ⱖ 0
In Exercises 21–28, use the simplex method to maximize the In Exercises 31–36, solve the minimization problem by solving the
objective function, subject to the given constraints. dual maximization problem by the simplex method.
z ⫽ x1 ⫹ 2x2 z ⫽ 5x1 ⫹ 4x2 w ⫽ 9x1 ⫹ 15x2 w ⫽ 16x1 ⫹ 18x2
2x1 ⫹ x2 ⱕ 31 x1 ⫺ x2 ⱕ 22 x1 ⫹ 5x2 ⱖ 15 2x1 ⫺ 3x2 ⱖ 14
x1 ⫹ 4x2 ⱕ 40 x1 ⫹ 2x2 ⱕ 43 4x1 ⫺ 10x2 ⱖ 0 ⫺4x1 ⫹ 9x2 ⱖ 8
x1, x2 ⱖ 0 x1, x2 ⱖ 0 x1, x2 ⱖ 0 x1, x2 ⱖ 0
z ⫽ x1 ⫹ 2x2 ⫹ x3 z ⫽ 4x1 ⫹ 5x2 ⫹ 6x3 w ⫽ 24x1 ⫹ 22x2 ⫹ 18x3 w ⫽ 32x1 ⫹ 36x2 ⫹ 4x3
2x1 ⫹ 2x2 ⫹ x3 ⱕ 20 4x1 ⫹ 2x2 ⫹ x3 ⱕ 30 2x1 ⫹ 2x2 ⫺ 3x3 ⱖ 24 4x1 ⫹ 3x2 ⫺ x3 ⱖ 8
x1 ⫹ x2 ⫺ 2x3 ⱕ 23 x1 ⫹ 3x2 ⫹ 2x3 ⱕ 54 6x1 ⫺ 2x3 ⱖ 21 ⫺8x1 ⫹ 6x2 ⫺ 6x3 ⱖ 0
⫺2x1 ⫹ x2 ⫺ 2x3 ⱕ 8 x1 ⫹ x2 ⫹ 2x3 ⱕ 24 ⫺8x1 ⫺ 4x2 ⫹ 8x3 ⱖ 12 ⫺4x1 ⫹ 9x3 ⱖ 4
x1, x2, x3 ⱖ 0 x1, x2, x3 ⱖ 0 x1, x2, x3 ⱖ 0 x1, x2, x3 ⱖ 0
z ⫽ x1 ⫹ x2 z ⫽ 6x1 ⫹ 8x2 w ⫽ 16x1 ⫹ 54x2 ⫹ 48x3 w ⫽ 22x1 ⫹ 27x2 ⫹ 18x3
3x1 ⫹ x2 ⱕ 432 20x1 ⫹ 40x2 ⱕ 200 x1 ⫹ 2x2 ⫹ 3x3 ⱖ 2 ⫺2x1 ⫺ 3x2 ⫹ 6x3 ⱖ 0
x1 ⫹ 4x2 ⱕ 628 30x1 ⫹ 42x2 ⱕ 228 2x1 ⫹ 7x2 ⫹ 4x3 ⱖ 5 ⫺2x1 ⫹ 7x2 ⫹ 3x3 ⱖ 4
x1, x2 ⱖ 0 x1, x2 ⱖ 0 x1 ⫹ 3x2 ⫹ 4x3 ⱖ 1 2x1 ⫹ x2 ⫺ 3x3 ⱖ 12
27. Objective function: 28. Objective function: x1, x2, x3, ⱖ 0 x1, x2, x3 ⱖ 0
z ⫽ 3x1 ⫹ 5x2 ⫹ 4x3 z ⫽ 2x1 ⫹ 5x2 ⫺ x3

6x1 ⫺ 2x2 ⫹ 3x3 ⱕ 24 ⫺x1 ⫹ 3x2 ⫹ 2x3 ⱕ 92
3x1 ⫺ 3x2 ⫹ 9x3 ⱕ 33 ⫺2x1 ⫹ 2x2 ⫹ 12x3 ⱕ 76
⫺2x1 ⫹ x2 ⫺ 2x3 ⱕ 25 3x1 ⫹ x2 ⫺ 6x3 ⱕ 24
x1, x2, x3 ⱖ 0 x1, x2, x3 ⱖ 0
332600_09_5.qxp 4/22/08 4:53 PM Page 593
Chapter 9 Review E xercises 593
In Exercises 37–42, use the simplex method to solve the mixed- 45. A company owns three mines that have the daily production
constraint problem. levels (in metric tons) shown in the table.
37. (Maximize) 38. (Maximize) Grade of Ore
Objective function: Objective function: Mine High Medium Low
z ⫽ x1 ⫹ 2x2 z ⫽ 2x1 ⫹ 3x2
A 1 2 3
⫺4x1 ⫹ 2x2 ⱕ 26 ⫺x1 ⫹ x2 ⱖ 40 B 1 2 2
⫺3x1 ⫹ x2 ⱖ 12 x2 ⱕ 61 C 2 1 1
x1, x2 ⱖ 0 x1, x2 ⱖ 0
The company needs 60 metric tons of high-grade ore, 48 metric
39. (Maximize) 40. (Maximize) tons of medium-grade ore, and 55 metric tons of low-grade ore.
Objective function: Objective function: How many days should each mine be operated in order to
z ⫽ 2x1 ⫹ x2 ⫹ x3 z ⫽ 3x1 ⫹ 2x2 ⫹ x3 minimize the cost of meeting these requirements if the daily
operating costs are $200 for mine A, $200 for mine B, and $100
Constraints: Constraints: for mine C, and what would be the minimum total cost?
x1 ⫹ x2 ⫹ x3 ⱕ 60 2x1 ⫹ x2 ⫹ 3x3 ⱕ 52 46. Rework Exercise 45 using the daily production schedule shown
⫺4x1 ⫹ 2x2 ⫹ x3 ⱖ 52 x1 ⫹ x2 ⫹ 2x3 ⱖ 24 in the table.
2x1 ⫹ x3 ⱖ 40 2x2 ⫹ x3 ⱕ 52
Grade of Ore
x1, x2, x3 ⱖ 0 x1, x2, x3 ⱖ 0
Mine High Medium Low
41. (Minimize) 42. (Minimize)
A 2 1 2
z ⫽ 9x1 ⫹ 4x2 ⫹ 10x3 z ⫽ 4x1 ⫺ 2x2 ⫺ x3 B 1 1 1
Constraints: Constraints: C 4 2 1
32x1 ⫹ 16x2 ⫹ 8x3 ⱕ 344 2x1 ⫺ x2 ⫺ x3 ⱕ 41
The company needs 190 metric tons of high-grade ore, 120
20x1 ⫺ 40x2 ⫹ 20x3 ⱖ 200 x1 ⫺ 2x2 ⫺ x3 ⱖ 10 metric tons of medium-grade ore, and 200 tons of low-grade ore.
⫺45x1 ⫹ 15x2 ⫹ 30x3 ⱕ 525 ⫺x1 ⫺ 7x2 ⫹ 5x3 ⱕ 11 The daily operating costs are $200 for mine A, $150 for mine B,
x1, x2, x3 ⱖ 0 x1, x2, x3 ⱖ 0 and $125 for mine C.
43. A tailor has 12 square feet of cotton, 21 square feet of silk, and True or False? In Exercises 47–50, determine whether each state-
11 square feet of wool. A vest requires 1 square foot of cotton, ment is true or false. If a statement is true, give a reason or cite an
2 square feet of silk, and 3 square feet of wool. A purse requires appropriate statement from the text. If a statement is false, provide
2 square feet of cotton, 1 square foot of silk, and 1 square foot an example that shows the statement is not true in all cases or cite
of wool. If the purse sells for $80 and the vest sells for $50, how an appropriate statement from the text.
many purses and vests should be made to maximize the tailor’s
profit? What is the maximum revenue? 47. A linear programming problem is in standard form if it seeks to
44. A traditional wood carpentry workshop has 400 board-feet of minimize the objective function subject to constraints.
plywood, 487 board-feet of birch, and 795 board-feet of pine. A 48. To locate the entering column, use the greatest positive value in
wooden bar stool requires 1 board-foot of plywood, 2 board-feet the bottom row.
of birch, and 1 board-foot of pine. A wooden step stool requires 49. In the final tableau, the maximum solution is the entry located in
1 board-foot of plywood, 1 board-foot of birch, and 3 board-feet the lower right corner of the tableau.
of pine. A wooden ottoman requires 2 board-feet of plywood, 50. To solve a minimization problem in standard form is equivalent
1 board-foot of birch, and 1 board-foot of pine. If the bar stool to solving the dual maximization problem.
sells for $22, the step stool sells for $42, and the ottoman
sells for $29, what combination of products would yield the
maximum gross income?
332600_09_5.qxp 4/22/08 4:53 PM Page 594
CHAPTER 9 Project
Cholesterol in human blood is necessary, but too much can lead to health problems.
A blood test gives three readings, LDL (“bad”) cholesterol, HDL (“good”) choles-
terol, and total cholesterol. It is recommended that your LDL cholesterol level be
less than 130 milligrams per deciliter, your HDL cholesterol level be at least 40
milligrams per deciliter, and your total cholesterol level be no more than 200
milligrams per deciliter.
Write a system of linear inequalities for the recommended cholesterol levels. Let x
represent the HDL cholesterol level, and let y represent the LDL cholesterol level.
Graph the system of linear inequalities. Find and label all vertices.
Part I
Determine if the cholesterol levels listed below are within recommendations. Explain
your reasoning.
LDL: 120 milligrams per deciliter
HDL: 90 milligrams per deciliter
Total: 210 milligrams per deciliter
Give an example of cholesterol levels in which the LDL cholesterol level is too high
but the HDL cholesterol level is acceptable.
Another recommendation is that the ratio of total cholesterol to HDL cholesterol be
less than 4 (that is, less than 4 to 1). Find a point in the solution region of your
system of inequalities that meets this recommendation. Explain why it meets the
recommendation.
Part II
Ask a friend or relative what his or her cholesterol levels are (or use your own). See
if the levels are within each acceptable range. How far below or above the accepted
levels are the results? Compare your results with those of other students in the class.

Linear Programming Chapter: Systems of Linear Inequalities

Încărcat de

Informații document

Descriere originală:

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Linear Programming Chapter: Systems of Linear Inequalities

Încărcat de

Drepturi de autor:

Formate disponibile

332600_09_1.

qxp 4/22/08 4:50 PM Page 531

9.1 Systems of Linear CHAPTER OBJECTIVES

9.1 Systems of Linear Inequalities

532 Chapter 9 Linear Programming

EXAMPLE 1 Sketching the Graph of a Linear Inequality

Figure 9.1 Figure 9.2

Section 9.1 Sy stems of Linear Inequalities 533

EXAMPLE 2 Sketching the Graph of a Linear Inequality

534 Chapter 9 Linear Programming

A solution of a system of inequalities in x and y is a point 共x, y兲 that satisfies each

EXAMPLE 3 Solving a System of Inequalities

Section 9.1 Sy stems of Linear Inequalities 535

Border lines can intersect x + y < −1

536 Chapter 9 Linear Programming

EXAMPLE 4 An Application of a System of Inequalities

Section 9.1 Sy stems of Linear Inequalities 537

For vitamin C: 10x ⫹ 30y ⱖ 90

SECTION 9.1 Exercises

538 Chapter 9 Linear Programming

(e) y (f) y 33. ⫺3x ⫹ 2y < 6 34. x ⫺ 7y > ⫺36

37. 2x ⫹ y > 2 38. x ⫺ 7y > ⫺36

Section 9.1 Sy stems of Linear Inequalities 539

540 Chapter 9 Linear Programming

9.2 Linear Programming Involving Two Variables

EXAMPLE 1 Solving a Linear Programming Problem

subject to the constraints listed below.

Section 9.2 Linear Programming Involving Two Variables 541

542 Chapter 9 Linear Programming

These guidelines will work regardless of whether the objective function is to be

EXAMPLE 2 Solving a Linear Programming Problem

where x ⱖ 0 and y ⱖ 0, subject to the constraints

Section 9.2 Linear Programming Involving Two Variables 543

Figure 9.13 Figure 9.14

EXAMPLE 3 Minimizing an Objective Function

where x ⱖ 0 and y ⱖ 0, subject to the constraints

544 Chapter 9 Linear Programming

EXAMPLE 4 An Unbounded Region

where x ⱖ 0 and y ⱖ 0, subject to the constraints

Section 9.2 Linear Programming Involving Two Variables 545

EXAMPLE 5 An Application: Minimum Cost

546 Chapter 9 Linear Programming

SECTION 9.2 Exercises

Section 9.2 Linear Programming Involving Two Variables 547

548 Chapter 9 Linear Programming

41. If an objective function has a maximum value at the vertices

550 Chapter 9 Linear Programming

9.3 The Simplex Method: Maximization

Section 9.3 The Simplex Method: Maximization 551

A basic solution of a linear programming problem in standard form is a solution

The Simplex Tableau

552 Chapter 9 Linear Programming

This solution is a basic feasible solution and is often written as

EXAMPLE 1 Pivoting to Find an Improved Solution

The objective function for this problem is z ⫽ 4x1 ⫹ 6x2.

Section 9.3 The Simplex Method: Maximization 553

The new tableau now appears as follows.

554 Chapter 9 Linear Programming

So, the new simplex tableau is as follows.

Section 9.3 The Simplex Method: Maximization 555

556 Chapter 9 Linear Programming

Note that the basic feasible solution of an initial simplex tableau is