LINEAR PROGRAMMING PROBLEM
INTRODUCTION
Why do we study Linear Programming (LP)?
Many of our real life problems can be converted into mathematical problems. Say for
example:
• Maximizing the profit of an industry
• Minimizing our monthly budget
• Minimizing the cost of the production
az
and many more...
So to solve such type of problems, we study LP.
How to convert real life problems in LP?
(I) Identify the problem
N
(II) Translate it into mathematical terms/model
ba
(III) Find the solution that satisfy the model
(IV) Translate the result obtained back to its original real life form
Area Problem (Non-Linear)
di
■ Example 1 Maximize the area of the rectangle, if the perimeter of the same rectangle
is 100 units.
Solution Let x be the height and y be the breadth of the rectangle. Then by the second
A
statement, we get 2(x + y) = 100. Also, we must have x, y ≥ 0. Let z be the area of the
rectangle, then according to the question, we need to have
Maximize z = xy
Mathematical Model of Area Problem:
Maximize z = xy → Objective Function
subject to
)
2(x + y) = 100
Constraints
x, y ≥ 0
2
Definition 1 The constraint x, y ≥ 0 is called as non-negative constraint and the
variables x, y are called as decision variables.
Definition 2 — Feasible Solution. A solution that satisfies all the constraints
(including the non-negative constraint) of the LP is called as feasible solution.
Otherwise the solution is called as the infeasible solution.
Definition 3 — Optimal Solution. A feasible solution that gives the maximum or
minimum value of the objective function is called as optimal solution.
az
Methods to Find the Solution of a LP
There are three methods to find the solution of a LP:
N
• Graphical Method (only for 2 variables)
• Basic Feasible Solution (for n variables)
• Simplex Method (for n variables)
ba
GRAPHICAL METHOD
■ Example 2 — Maximization Problem. An industry produces both interior and
exterior paints from two raw materials, M1 and M2. The following table provides the basic
di
data of the problem:
A
The daily demand for interior paint cannot exceed that for exterior paint by more than 1
ton. Also, the maximum daily demand for interior paint is 2 tons. Determine the optimum
(best) product mix of interior and exterior paints that maximizes the total daily profit.
Solution Let
x1 = tons produced daily of exterior paint
x2 = tons produced daily of interior paint
3
Then the objective function becomes
Maximize z = 5x1 + 4x2
and the constraints are
6x1 + 4x2 ≤ 24 (Raw Material M1) (1)
x1 + 2x2 ≤ 6 (Raw Material M2) (2)
x2 ≤ x1 + 1 =⇒ −x1 + x2 ≤ 1 (Market Limit) (3)
x2 ≤ 2 (Demand Limit) (4)
x1 ≥ 0 (Non Negative Constraint) (5)
az
x2 ≥ 0 (Non Negative Constraint) (6)
• x1 ≥ 0 and x2 ≥ 0 prove that the feasible solution exists in first quadrant.
N
• Replace ≥ or ≤ by = in equations (1) to (4), that is,
6x1 + 4x2 = 24
x1 + 2x2 = 6
ba
−x1 + x2 = 1
x2 = 2
Then plot these lines on the graph.
di
A
4
Feasible Points Value of z = 5x1 + 4x2
A(0,0) 0
B(4,0) 20
C(3,1.5) 21→ Maximum
D(2,2) 18
E(1,2) 13
F(0,1) 4
az
Therefore the solution is C(3,1.5). This means that 3 tons are produced of exterior paint
and 1.5 tons are produced of interior paint daily to make the daily profit of $21,000
■ N
Example 3 — Minimization Problem. To feed her stock a farmer can purchase two
kinds of feed. The farmer has determined that the herd requires 60, 84, and 72 units of the
nutritional elements A, B, and C, respectively, per day. The contents and cost of a pound
of each of the two feeds are given in the following table.
ba
di
A
Determine the least expensive way of providing an adequate diet by combining the two
feeds.
Solution Let
x = lb of Feed 1
y = lb of Feed 2
Then the objective function becomes
Minimize z = 10x + 4y
5
and the constraints are
3x + 2y ≥ 60 (For A)
7x + 2y ≥ 84 (For B)
3x + 6y ≥ 72 (For C)
x, y ≥ 0 (Non Negative Constraint)
az
N
ba
di
Feasible Points Value of z = 10x + 4y
A
(0,42) 168
(6,21) 144→ Minimum
(18,3) 192
(24,0) 240
Therefore the solution is (6,21). This means that 6 lb of Feed 1 and 21 lb of Feed 2 are
combine together to provide an adequate diet at the cost of 144 cents.
■ Example 4 — Infinite Solution. Minimize z = x1 + x2
6
subject to
5x1 + 9x2 ≤ 45
x1 + x2 ≥ 2
x2 ≤ 4
x1 , x2 ≥ 0.
Solution The graph for the give LPP is:
az
N
ba
di
Feasible Points Value of z = x1 + x2
(9,0) 9
A
(2,0) 2→ Minimum
(0,4) 4
(0,2) 2→ Minimum
9 29
,4
5 5
This means all the points lying on the line joining (0, 2) and (2, 0), that is, x1 + x2 = 2
will minimize z as the value of the objective function is same at these two points. Therefore
infinite number of solutions exist for this LP.
■ Example 5 — No Feasible Solution. Maximize z = 3x1 − 2x2
7
subject to
2x1 + x2 ≤ 2
3x1 + 4x2 ≥ 12
x1 , x2 ≥ 0.
Solution The graph for the given LPP is:
az
N
ba
di
A
Since there is no common region, therefore no feasible solution exists for this LP.
■ Example 6 — Unbounded Solution. Maximize z = 3x1 + 4x2
subject to
−3x1 + 2x2 ≤ 6
−x1 + 3x2 ≤ 18
x1 , x2 ≥ 0.
Solution The graph for the given LPP is:
8
az
Here the feasible region is unbounded.
N
ba
Feasible Points Value of z = 3x1 + 4x2
di
(0,0) 0
(0,3) 12
18 48 246
A
, → Maximum
7 7 7
Since the region is unbounded, 246/7 may or may not be the maximum value of z. So
consider the region drawn by
246
3x1 + 4x2 >
7
Substituting (0, 0) will give 0 > 246/7 which is false. We see that the resulting open half
plane has points in common with feasible region. Therefore 246/7 is not the maximum
value. Thus in this case, we say optimal solution is unbounded.
9
az
Points to Remember N
Let the objective function be z = ax + by. Suppose M is the largest value and m is the
smallest value of z.
ba
• If the region is bounded, both M is the maximum value and m is the minimum value
of z.
• If the region is unbounded, then we
di
(a) M is the maximum value of z, if the open half plane determined by ax + by > M
has no point in common with the feasible region. Otherwise, z has no maximum
value.
(b) Similarly, m is the minimum value of z, if the open half plane determined by
A
ax + by < m has no point in common with the feasible region. Otherwise, z has
no minimum value.
