You are on page 1of 9

BUSI2013 Decision Analysis

Solving LP Problems using the Graphical Solution Method.

We are given the following three LP models.

Problem 1:1
 2 + 3

. .
1 + 2 6
5 + 3 15
, 0

Problem 2:
 3 + 1

. .
1 + 2 6
5 + 3 15
, 0

Problem 3:
 1 + 3

. .
1 + 2 6
5 + 3 15
, 0

Note that the constraints for all three problems are identical. The only difference between models is in the
objective function.

Problem 1 corresponds to Chapter 7 Question 10 on the Unit 3 assignment.

Graphing Constraints & Identifying the Feasible Region


Because the constraints are identical for all three problems, lets start by plotting the constraints, locating the
feasible region, and identifying the extreme points.

Every possible solution satisfying all the constraints is located within the feasible region. There are four
extreme points in this feasible region. They are located where:
1)   0 (or the B-axis) intersects 1 + 2  6,
2) 1 + 2  6 intersects 5 + 3  15,
3) 5 + 3  15 intersects  0 (or the A-axis), and
4)   0 intersects  0 (or the origin).
 
,


The four extreme points are, therefore, located at the following , -points respectively: 0,3,   ,
and 0,0. See the math for calculating these points in Appendix A at the end of the document.

3,0,

These four extreme points are common across all three problems.

Finding the Optimal Solutions


Note that the constraints, on their own, only give us information on the range of possible solutions that satisfy
the constraints. We do not yet have enough information to determine the optimal solution. To find the optimal
solution we have to use information from the objective function.

Problem 1
Lets start by finding the optimal solution for Problem 1. There are two methods we can use to find this:
1) Calculating the objective values for each of the extreme points,
2) Using test lines for the objective function.

Finding Optimal Solution using Objective Values of the Extreme Points


We substitute the extreme points into the objective function, 2 + 3 , to calculate the objective values at each
point.
Extreme Point

Objective Value

,   0,3

12 15
,    ,
7 7
,   3,0
,   0,0

69
6
 9
7
7
6
0

Since this is a maximization problem, we identify the extreme point with the largest objective value as the
 



optimal solution. In this case, the table above allows us to identify ,     ,
solution with an optimal value of

!"


 9 .

 1  , 2  as the optimal

Finding Optimal Solution using Objective Values of the Extreme Points


We first choose a series of values for the objective function, and plot these values over the constraint graph.
In this case let us plot the following two test lines:
1) 2 + 3  9, and


2) 2 + 3  9 

We are looking to maximize the objective function, so we want the largest objective value. We see that as we


increase the value of the objective function from 9 to 9 , the number of values on the objective function line


that satisfy the constraints decrease (less of the line segment is within the feasible region). We can also see
visually that if we increase the value of the objective function a bit more we can find a value for which the
objective function test line will touch the feasible region at only one point: the intersection of 1 + 2  6 and
5 + 3  15. This indicates that this point is the optimal solution.

We have solved for the intersection of these constraints above. The optimal solution is located at the point
,   

 
, 
 

with an objective value of

!"


 9 .

Problem 2
We know find the optimal solution for Problem 2 using both methods.
Finding Optimal Solution using Objective Values of the Extreme Points
We substitute the extreme points into the objective function, 3 + 1 , to calculate the objective values at each
point.
Extreme Point

Objective Value

,   0,3

12 15
,    ,
7 7
,   3,0
,   0,0

51
2
7
7
7
9
0

Since this is a maximization problem, we identify the extreme point with the largest objective value as the
optimal solution. In this case, the table above allows us to identify ,   3,0 as the optimal solution with
an optimal value of 9

Finding Optimal Solution using Objective Values of the Extreme Points


We first choose a series of values for the objective function, and plot these values over the constraint graph.
In this case let us plot the following two test lines:
1) 3 + 1  8, and
2) 3 + 1  9

We are looking to maximize the objective function, so we want the largest objective value. We see that as we
increase the value of the objective function from 8 to 9, the number of values on the objective function line that
satisfy the constraints decrease (less of the line segment is within the feasible region). We can also see
visually that if we increase the value of the objective function a bit more we can find a value for which the
objective function test line will touch the feasible region at only one point: the intersection of  0 and
5 + 3  15. This indicates that this point is the optimal solution.
We have solved for the intersection of these constraints above. The optimal solution is located at the point
,   3.0 with an objective value of 9

Problem 3
We know find the optimal solution for Problem 2 using both methods.

Finding Optimal Solution using Objective Values of the Extreme Points


We substitute the extreme points into the objective function, 1 + 3 , to calculate the objective values at each
point.
Extreme Point

Objective Value

,   0,3

12 15
,    ,
7 7
,   3,0
,   0,0

57
1
 8
7
7
3
0

Since this is a maximization problem, we identify the extreme point with the largest objective value as the
optimal solution. In this case, the table above allows us to identify ,   0,3 as the optimal solution with
an optimal value of 9

Finding Optimal Solution using Objective Values of the Extreme Points


We first choose a series of values for the objective function, and plot these values over the constraint graph.
In this case let us plot the following two test lines:
1) 1 + 3  8, and
2) 1 + 3  9

We are looking to maximize the objective function, so we want the largest objective value. We see that as we
increase the value of the objective function from 8 to 9, the number of values on the objective function line that
satisfy the constraints decrease (less of the line segment is within the feasible region). We can also see
visually that if we increase the value of the objective function a bit more we can find a value for which the
objective function test line will touch the feasible region at only one point: the intersection of   0 and
1 + 2  6. This indicates that this point is the optimal solution.
We have solved for the intersection of these constraints above. The optimal solution is located at the point
,   0,3 with an objective value of 9.

Conclusions
We have found the optimal solutions to all three of the problems using two variations of the graphical solution
method.
We have seen that despite having the same constraints (and therefore the same feasible regions and extreme
points), the three problems had different optimal solutions. There was not enough information in the
constraints alone to find the optimal solution. We had to make use of information (and therefore refer to the
information) in the objective function to find the solution.
In summary, the graphical solution method requires at least three items for a complete solution:
1) A plot / graph of the constraints (using information in the constraints, we can determine the feasible
region and extreme points),
2) An argument using the objective function (either through the objective values at each of the extreme
points or using test lines) to determine the approximate location of the optimal solution, and
3) Determination of the optimal solution (say, calculating the intersection of the constraints defining the
optimal solution).

Appendix A Identification of the extreme points


We identified earlier that the four extreme points are located where:
1)   0 (or the B-axis) intersects 1 + 2  6,
2) 1 + 2  6 intersects 5 + 3  15,
3) 5 + 3  15 intersects  0 (or the A-axis), and
4)   0 intersects  0 (or the origin).

Extreme Pont 1:
Located at the intersection of   0 and 1 + 2  6
We know   0, we substitute this into the second equation to get
1 + 2  6
10 + 2  6
0 + 2  6
2  6
6

2
3
So the first extreme point is at ,   0,3

Extreme Pont 2:
Located at the intersection of 1 + 2  6 and 5 + 3  15.
To find the intersection we want to isolate one of the variables. Lets start by multiplying the first equation by 5
and subtracting the second equation:
5 + 10  30
5 + 3  15
------------------------------7  15
15

7
Substitute this into the first equation,
1 + 2  6
15
1 + 2 
 6
7

30
 6
7
30
1  6
7
42 30
1 

7
7
12

7
1 +

So the extreme point is at ,   

 
, 
 

Extreme Pont 3:
Located at the intersection of 5 + 3  15 and  0 (or the A-axis).
Substitute  0 into the first equation:
5 + 3  15
5 + 30  15
5 + 0  15
5  15
15

5
3
So the extreme point is at ,   3,0

Extreme Pont 4:
Located at the intersection of   0 and  0, or ,   0,0

The four extreme points are (written in the form , ) are:

0,3, 

 
, ,
 

3,0, and 0,0.

You might also like