"Just The Maths" Unit Number 5.4 Geometry 4 (Elementary Linear Programming) by A.J.Hobson
"Just The Maths" Unit Number 5.4 Geometry 4 (Elementary Linear Programming) by A.J.Hobson
UNIT NUMBER
5.4
GEOMETRY 4
(Elementary linear programming)
by
A.J.Hobson
(i) The equation, y = mx + c, of a straight line is satisfied only by points which lie on the
line. But it is useful to investigate the conditions under which a point with co-ordinates
(x, y) may lie on one side of the line or the other.
(ii) For example, the inequality y < mx + c is satisfied by points which lie below the line
and the inequality y > mx + c is satisfied by points which lie above the line.
(iii) Linear inequalities of the form Ax+By +C < 0 or Ax+By +C > 0 may be interpretted
in the same way by converting, if necessary, to one of the forms in (ii).
(v) Several simultaneous linear inequalties may be used to determine a region of the xy-
plane throughout which all of the inequalities are satisfied. The region is called the “feasible
region”.
EXAMPLES
x ≥ 0, y ≥ 0, x + y ≤ 20, and 3x + 2y ≤ 48
Solution
We require the points of the first quadrant which lie on or below the straight line
y = 20 − x and on or below the straight line y = − 23 x + 16.
The feasible region is shown as the shaded area in the following diagram:
1
y
6
24
20
0 - x
16 20
Solution
We require the points which lie on or to the left of the straight line x = 50, on or below
the straight line y = 30, on or below the straight line y = 80 − x and on or above the
straight line y = 60 − x.
The feasible region is shown as the shaded area in the following diagram:
y
6
80
60
30
10
0 30 50 60 80
- x
2
5.4.2 OBJECTIVE FUNCTIONS
An important application of the feasible region discussed in the previous section is that
of maximising (or minimising) a linear function of the form px + qy subject to a set of
simultaneous linear inequalities. Such a function is known as an “objective function”
Essentially, it is required that a straight line with gradient − pq is moved across the appropriate
feasible region until it reaches the highest possible point of that region for a maximum value
or the lowest possible point for a minimum value. This will imply that the straight line
px + qy = r is such that r is the optimum value required.
However, for convenience, it may be shown that the optimum value of the objective function
always occurs at one of the corners of the feasible region so that we simply evaluate it at
each corner and choose the maximum (or minimum) value.
EXAMPLES
1. A farmer wishes to buy a number of cows and sheep. Cows cost £18 each and sheep
cost £12 each.
The farmer has accommodation for not more than 20 animals, and cannot afford to
pay more than £288.
If he can reasonably expect to make a profit of £11 per cow and £9 per sheep, how
many of each should he buy in order to make his total profit as large as possible ?
Solution
Suppose he needs to buy x cows and y sheep; then, his profit is the objective function
P ≡ 11x + 9y.
Also,
Thus, we require to maximize P ≡ 11x + 9y in the feasible region for the first example
of the previous section.
The corners of the region are the points (0, 0), (16, 0), (0, 20) and (8, 12), the last of
these being the point of intersection of the two straight lines x+y = 20 and 3x+2y = 48.
The maximum value occurs at the point (8, 12) and is equal to 88 + 108 = 196. Hence,
the farmer should buy 8 cows and 12 sheep.
3
2. A cement manufacturer has two depots, D1 and D2 , which contain current stocks of 80
tons and 20 tons of cement respectively.
Two customers C1 and C2 place orders for 50 and 30 tons respectively.
The transport cost is £1 per ton, per mile and the distances, in miles, between D1 , D2 ,
C1 and C2 are given by the following table:
C1 C2
D1 40 30
D2 10 20
From which depots should the orders be dispatched in order to minimise the transport
costs ?
Solution
Suppose that D1 distributes x tons to C1 and y tons to C2 ; then D2 must distribute
50 − x tons to C1 and 30 − y tons to C2 .
All quantities are positive and the following inequalties must be satisfied:
The total transport costs, T , are made up of 40x, 30y, 10(50 − x) and 20(30 − y).
That is,
4
5.4.3 EXERCISES
1. Sketch, on separate diagrams, the regions of the xy-plane which correspond to the
following inequalities (assuming that x ≥ 0 and y ≥ 0):
(a)
x + y ≤ 6;
(b)
x + y ≥ 4;
(c)
3x + y ≥ 6;
(d)
x + 3y ≥ 6.
2. Sketch the feasible region for which all the inequalities in question 1 are satisfied.
3. Maximise the objective function 5x + 7y subject to the simultaneous linear inequalities
x ≥ 0, y ≥ 0, 3x + 2y ≥ 6 and x + y ≤ 4.
How many shifts per week should each seam be worked, in order to fulfill the contracts
most economically ?
5
5. A manufacturer employs 5 skilled and 10 semi-skilled workers to make an article in two
qualities, standard and deluxe.
The deluxe model requires 2 hour’s work by skilled workers; the standard model requires
1 hour’s work by skilled workers and 3 hour’s work by semi-skilled workers.
No worker works more than 8 hours per day and profit is £10 on the deluxe model and
£8 on the standard model.
How many of each type, per day, should be made in order to maximise profits ?
y
6
6
O -x
6
6
(b) The region is as follows:
y
6
O -x
4
y
6
6
O -x
2
7
(d) The region is as follows:
y
6
O -x
6
y
6
6
O -x
2 4 6
8
3. The feasible region is as follows:
y
6
4
O -x
2 4
The maximum value of 5x + 7y occurs at the point (0, 4) and is equal to 28.
4. Subject to the simultaneous inequalities
the function 2x + 5y has minimum value 20 at any point on the line 2x + 5y = 20.
5. Subject to the simultaneous inequalities
x ≥ 0, y ≥ 0, x + 2y ≤ 0 and 3x + 2y ≤ 80,
the objective function P ≡ 8x + 10y has maximum value 260 at the point (20, 10).