Sei sulla pagina 1di 23

# LINEAR PROGRAMMING

Graphical Solution
Reddy Mikks Company
• The Reddy Mikks produces both interior and exterior paints from
two raw materials, M1 and M2. The following table provides the
basic data of the problem:
Maximum daily
Tons of raw material per ton of
availability
Exterior Paint Interior Paint (tons)
Raw material, M1 6 4 24
Raw Material, M2 1 2 6
Profit per ton (\$1000) 5 4
• A market survey restricts the maximum daily demand of interior
paint to 2 tons. Additionally the daily demand for interior paint
cannot exceed that of exterior paint by more than 1 ton. Reddy
Mikks wants to determine the optimum (best) product mix of interior
and exterior paints that maximize the total daily profit.
• Final Formulation

## • Now the complete Reddy Mikks model is written as

• Maximise z = 5x1 + 4x2
• Subject to 6x1 + 4x2 ≤ 24
• x1 + 2x2 ≤ 6
• -x1 + x2 ≤ 1
• x2 ≤ 2
• x1 ≥0
x2 ≥ 0
• Any solution that satisfies all the constraints of the model is a
feasible solution.
Tentative Solution to Reddy Mikks
• Solution x1 = 3 and x2 = 1 tons is feasible because it doesn’t
violate the constraints including non negative restriction.

## • First constraint 6x1 + 4x2 ≤ 24 → 6 x 3 + 4 x 1 ≤ 24 → 22 ≤ 24

• Second constraint x1 + 2x2 ≤ 6 → 3 + 2 x 1 ≤ 6 → 5 ≤ 6
• Third constraint -x1 + x2 ≤ 1 → -3 + 1 ≤ 1 → -2 ≤ 1
• Fourth constraint x2 ≤ 2 or 1 ≤ 2
• Finally x1, x2 ≥ 0
• Value of objective function is z = 5 x1 + 4x2 = 5 x 3 + 4 x 1 = 19
(\$1000)

## • In this way we get large number of feasible solution of the above

model, but our aim is to determine optimum feasible solution
that yields maximum total profit, which is impossible to use
direct substitutions to determine the optimum.
Solution of Linear Programming

## • After formulation of a given situation as a linear

programming problem, one has to solve the problem for
finding the variables that optimize the objective function.

## • Algebraic Algorithm (Simplex Method) for two and more

than two variables

## • Optimization Software (TORA,CMMS, AMPL, Excel

Solver, etc) for two and more than two variables
Graphical Method of Solution

## 1. Determination of the solution space which define the feasible

solutions that satisfy all the constraints of the model.

## 2. Determination of the optimal solution from among all the

points in the feasible solution space.

## • This describes for both a maximization and minimization

objective function.
• Example: Reddy Mikks Model

## • Step 1 Determination of the Solution Space:

• Let the horizontal axis x1 and the vertical axis x2 represent the
exterior paint and interior paint variables respectively.

## • Next, considering non negative restrictions x1 ≥ 0 and x2 ≥ 0.

• These two constraints restrict the solution space area to the first
quadrant which lie above x1 axis and right of the x2 axis.

## • By replacing inequalities in the constraint we get linear equation

and then plot the resulting straight line we get the graph.
• To plot the line we need two distinct point

## • Obtained x2 = 24/4 = 6 and x1 = 24/6 = 4.

• Thus the line passes through two points (0,6) and (4,0).

• Similarly, all other points are found from rest of the constraints

• All the inequalities divide (x1, x2) plane into two half spaces

• One side satisfies the inequality and other one does not.
• Procedure for determining the feasible side is to use the origin
(0, 0) as a reference point

• e.g. the first constraint i.e. 6x1 + 4x2 ≤ 24 satisfy the point (0, 0)
which yield 6 x 0 + 4 x 0 ≤ 24.

## • This means the feasible side of the constraint 6x1 + 4x2 ≤ 24

includes the origin is shown by arrow associated with the
constraint 1.

• If origin does not satisfy the inequality the direction arrow must
point in the opposite side of (0, 0).

## • Also if line happens to pass through origin, then we can choose

another reference point to effect the desire result.
• Step 2 Determination of the Optimum Solution:

## • Graph provides the feasible solution space which delineated by

the line segments joining the corner point ABCDE and F.

## • Any point within or on the boundary of the space ABCDEF is

feasible point as it satisfies all the constraints .

## • Optimal solution identifies the direction in which the profit

function z = 5x1+4x2 increases as shown in the graph.

• It is clear that the line 5x1+4x2 passing through that point move
up parallel to itself and thus increase its value.

## • One of the corner points (ABCDE and F) of the feasible region

will be optimal solution.
• Despite infinite number of feasible solutions one has to
consider a finite number of corner points to determine optimal
solution.

along the boundary line of feasible region to adjacent corner
point that gives the largest increase in the objective function.

## • Repeat this iteration until to reaches a corner point from which

it does not pay to move to any of its adjacent corner points.

## • This is the basis and logic of the simplex method which is

applied to the large size linear programming problems.

## • After move from A to B, then move from B to C (3, 3/2) profit

increase in 5x3 + 4x(3/2) = 21.

## • Now move from C to D (2, 2) profit will be 5x2 + 4x2 = 18, in

this direction profit decreases, so it do not move from C to D.

## • Hence, optimal solution at point C with x1 = 3 tons, x2 = 1.5 tons

of exterior and interior paint will yield a daily profit of \$21000.
Corner point (x1, x2) z
A (0,0) 0
B (4,0) 20
C (3,1.5) 21 Optimum
D (2,2) 18
E (1,2) 13
F (0,1) 4
Example: DIET PROBLEM

## • Ozark Farms uses at least 800lb of special feed daily. The

special feed is a mixture of corn and soybean meal with the
following composition:
lb per lb of feedstuff
Feedstuff Protein Fiber Costs(\$/lb)
Corn 0.09 0.02 0.30
Soybean meal 0.60 0.06 0.90

## The dietary requirements of the special feed stipulate at least

30% protein and at most 5% fibre. Ozerk Farms wishes to
determine the daily minimum cost feed mix.
Solution

## • Feed mix consists of corn and soybean meal, the decision

variables of the model are defined as

## • x1 = lb of corn in the daily mix

• x2 = lb of soybean meal in the daily mix

## • Objective function is to maximize total daily cost (in \$) of the

feed mix i.e. Minimize z = 0.3 x1 + 0.9 x2

## • Constraints must reflect the daily amount needed and the

dietary requirements.

## • Ozark Farms needs 800lb of feed a day, the associate

constraint can be expressed as x1 + x2 ≥ 800
• Protein dietary requirement constraint is developed next.

## • Amount of protein included in x1 lb of corn and x2 lb of

soybean meal is (0.09 x1 + 0.6 x2) lb.

• This quantity should equal at least 30% of the total feed mix
(x1 + x2) lb.

## • After rearranging we get the complete model

• The complete model thus becomes

## • Minimise z = 0.3 x1 + 0.9 x2

• Subject to x1 + x2 ≥ 800
• 0.21 x1 – 0.3 x2 ≤ 0
• 0.03 x1 – 0.01 x2 ≥ 0
• x1 ≥ 0
• x2 ≥ 0
The graphical solution is
• In graphical solution two constraints pass through origin.

## • The intersection of the line x1 + x2 ≥ 800 with the lines 0.21

x1 – 0.3 x2 ≤ 0 and 0.03 x1 – 0.01 x2 ≥ 0 and yield the
value of x1 and x2.

## • These two values put in the objective function and the

associate minimum cost is found to be \$ 473.64.
Slack Variables

## • Slack Variables: For constraints of the type (≤), the right

hand side normally represents the limit on the availability of a
resource and left hand side represents the usages of this limited
resource by the different variable of the model.
• A slack thus represents the amount by which the available
amount of resource exceeds its usage by the activities.

## • The constraint 6x1 + 4x2 ≤ 24 usage of material M1. It can be

equivalent to 6x1 + 4x2 + s1 = 24, provided that s1≥ 0.

## • The slack variable s1 = 24 - 6x1 - 4x2, thus represent the

unused amount of raw material M1.
Surplus and Unrestricted Variable

## • Surplus Variable: Constraint of the type (≥) normally set

minimum specification requirements.
• Surplus represents the excess of left hand side over the
minimum requirement.

## • In Diet Model: The constraint representing the minimum feed

requirements, x1+x2 ≥ 800 can be equivalent to x1 + x2 – s1 =
800 provided that s1 ≥ 0 surplus amount will be produced.

## • Unrestricted Variable: In Reddy Mikks and Diet models, the

nature of variables requires them to assume non negative real
values. There are situation where a variable can assume and
integer value.