Optimization Techniques: Linear
Programming
Linear Programming
Linear programming is a mathematical
technique
for
solving
constrained
maximization and minimization problems,
when there are many constraints and the
objective function to be optimized, as well
as the constraints faced, are linear (i.e.,
can be represented by straight lines).
It is a technique for providing specific
numerical solutions of problems.
It bridges the gap between abstract
economic theory and managerial decision
making in practice.
2
The
technique
was
developed
by
the
Russian mathematician L.V. Kantorovich in
1939
and
extended
by
the
American
mathematician G. B. Dantzig in 1947.
Use of linear programming
is expanding
very fast because of use of computer which
can
quickly
involving
the
solve
complex
optimal
use
problems
of
many
resources, which are available to a firm at a
particular time.
3
Difference between Traditional
Economic Analysis Vs. Linear
Programming
Both approaches show how economic
agents (producers or consumers) reach
optimal choices, how they do their planning
or programming in order to attain
maximum
utility,
maximum
profit,
minimum cost, etc.
Neither economic theory nor linear
programming say anything about the
implementation of the optimal plan or
solution. They simply derive the optimal
solution in any particular situation.
However,
In economic theory the optimal solution is
usually shown in qualitative abstract
terms, diagrams, or general mathematical
symbols. In contrast, linear programming
yields specific numerical solutions to the
particular optimization problems.
Relationships
of economic theory are
usually non-linear, expressed by curves
(not straight lines), while in linear
programming all relationships between
the variables involved are assumed to be
linear.
5
Suppose a firm has the following quantities of factors of
production
L= 400 units of labour (hours)
K= 300 units of capital (machine hours)
S = 1000 units of land (square feet)
The firm can produce either commodity x or commodity y
with the following available processes (activities)
activity A for x
activity B for y
Labour
lx = 4
ly = 1
Capital
Kx = 1
ky = 1
Land
Sx = 2
Sy = 5
The production of one unit of x requires 4 hours of labour,
1 machine hour and 2 square feet of land. Similarly, the
production of one unit of y requires 1 hour of labour, 1
machine hour and 5 square feet of land.
6
Suppose x yields a unit profit of INR 2, and commodity y
yields a unit profit of INR 1. The goal of the firm is to
choose the optimal product mix, that is, the combination
that maximises its total profit.
The total profit function can be written as
Z = 2X + 1Y
Where, Z = total profit; X = quantity of commodity x (or
level of Activity A); Y = quantity of commodity y (or level
of Activity B); and 2 and 1 are the unit profits of the two
commodities.
Objective Function: The total profit function is called the
objective function as it expresses the objective of the
firm. This is the function, which represents the goals of
the economic agent.
7
Constraints
Technical (or functional) constraints, and
Non-negativity constraints.
The technical constraints are set by the state of
technology and the availability of factors of
production.
There
are
many
technical
constraints as the factors of production
They express the fact that the quantities of factors
which will be absorbed in the production of the
commodities cannot exceed the available
quantities of these factors. Thus, in our problem
4 X 1Y 400;1X 1Y 300;2 X 5Y 1000
8
Where, X and Y are the levels of commodities x
and y, and integers on the left-hand side of the
equations are the technical coefficients of
production, i.e., the factors inputs required for the
production of one unit of the products x and y.
The figures on the right-hand side are the
resources that the firm has at its disposition.
The non-negativity constraints express the
necessity that the levels of production of the
commodities cannot be negative. The level of
production of any one commodity can either be
zero or positive, i.e.,
X 0;Y 0
The linear programming problem may be stated as
Maximize
Subject to
Z = 2X + 1Y
4 X 1Y 400;1X 1Y 300;2 X 5Y 1000
(technical constraints)
X 0; Y 0
(non-negativity constraints)
In linear programming the optimal solution is defined
by examining the set of possible alternatives solutions
and eliminating gradually the sub optimal ones until
the optimal is reached.
10
Graphical Solution of the Linear Programming Problem
Graphical
determination
of
the region of
feasible solutions,
Graphical
determination
of
the
objective
function, and
Determination of the optimal solution.
11
Graphical determination of the region of
feasible solutions
A solution will be feasible when it satisfies all the
constraints.
Here we have to satisfy both non-negativity
constraints as well as technical constraints.
Boundary set by the factor Labour: This is
defined by a straight line whose slope is the ratio
of the labour inputs in the production of the two
commodities.
Hence, slope of the line = input of L in x / input of L in y
= 4/1 = lx/ly
12
Similarly slope of the line for capital will be
= input of K in x/ input of K in y = Kx/Ky= 1/1=1
And, slope of the line for land will be
= input of S in x / input of S in y= Sx/ Sy = 2/5
As objective function may be represented by isoZ 2 X 1Y x X yY
profit lines i.e.,
Slope of the iso-profit line will be
Y
x unitprofit ofx 2
2
X
y unitprofit ofy 1
13
Determination of the Optimal Solution
The optimal solution will be found by the point of
tangency of the frontier of the region of feasible
solutions to the highest possible iso-profit curve.
The optimal solution will be a point on the
frontier of the region of all feasible solutions.
Here it will be Z = 2X + 1Y = 2(56) + 1 (178) = 290
14
The Duel Problem and Shadow Prices
The basic problem whose solution is attempted by
the linear programming technique is called the primal
problem.
To each primal problem corresponds a dual problem,
which yields additional information to the decisionmaker.
The nature of the dual problem depends on the
primal problem. If the primal problem is a
maximization problem, its dual is a minimization
problem.
From the solution of dual problem, we can derive the
shadow prices of the factors of production used by
the firm.
15
Shadow Prices
These are the imputed costs or opportunity costs
of the factors for a particular firm.
They are crucial indicators for the expansion of
the firm. They show which factors are
bottlenecks to the further expansion of the firm.
The shadow prices of the resources can be
compared with their market prices and help the
entrepreneur decide whether it is profitable to
hire additional units of these factors.
It decides how much the profit of the firm will be
increased if the firm employs an additional unit of
this factor.
16
Reference:
Koutsoyiannis,
Microeconomics,
London
A.
(1979),
Modern
Macmillan Press Ltd.,
17
