Linear Programming

Eng Mupfumira
Linear Programming
• First conceived by George B Dantzig around 1947
• The work of Kantorovich (1939) was published in 1959
• Dantzig’s first paper was titled ‘Programming Linear Structure’
• Koopmans coined the term ‘Linear Programming’ in 1948
• Simplex method was published in 1949 by Dantzig
Linear Programming
This technique is used to find a solution for optimising a given
objective. Objective may be maximizing profits or minimizing costs.
Objective function and Boundary conditions are linear in nature. LPP
techniques solve Product-Mix and Distribution problems of
enterprise. Its also used to allocate Scarce Resources in optimum
manner in problems of scheduling, product mix, etc.
Linear Programming Formulations
• Consider a small manufacturer making two products A and B.
• Two resources R1 and R2 are required to make these products
• Each unit of A requires 1 unit of R1 and 3 units of R2
• Each unit of B requires 1 unit of R1 and 2 units of R2
• The manufacturer has 5 units of R1 and 12 units of R2 available
• The manufacturer also makes a profit of
RTGS 6 per unit of product A sold and
RTGS 5 per unit of product B sold.
• The problem is that the manufacturer has to produce products A and
B such that profits are maximised
 Solution (Formulation)
• The manufacturer has to decide on the number of units of products A and
B to produce
• It is acceptable that the manufacturer would like to make as much profit as
possible and would decide on the production quantities accordingly
• The manufacturer has to ensure that the resources needed to make the
products are available
• Before we attempt to find out the decisions of the manufacturer, let us
redefine the problem in an algebraic form
• The manufacturer has to decide on the production quantities. Let us call
them x and y and define
• Let x be the number of units of product A made
• Let y be the number of units of product B made
Solution Continued
• The profit associated with x units of product A and y units of product
B is 6x + 5y. The manufacturer would determine x and y such that this
function has a maximum value.
• The requirements of the two resources are x + y for R1 and 3x + 2y for
R2 and the manufacturer has to ensure that these are available.
• The problem is to find x and y such that 6x + 5y is maximized and x + y
<= 5 and 3x + 2y <= 12 are satisfied. Also it is necessary that x,y >= 0
so that only nonnegative quantities are produced
• If we redefine the production quantities as x1 and x2 (for reasons of
uniformity and consistency) then the problem is to
 Terminology
• The problem variables x1 and x2 are called decision variables and they represent
solution or the output decision from the problem
• The profit function that the manufacturer wishes to increase, represents the
objective of making the decisions on the production quantities and is called the
objective function
• The conditions matching the resource availability and resource requirement are
called constraints
These usually limit the values the decision variables can take
• We have also explicitly stated that the decision variable should take nonnegative
values
This is true for all Linear Programming problems and is called nonnegativity restriction
• The problem that we have written down in algebraic form that represents the
mathematical model of the given system and is called the Problem Formulation.
• The problem formulation has the following steps
Identifying the decision variables
Writing the objective function
Writing the constraints
Writing the non negativity restrictions
• In the above formulation, the objective function and the constraints are
linear
Therefore the model that we formulated is a linear programming problem
• A Linear Programming problem has
A linear objective function
Linear constraints and
The non negativity constraints on all decision variables
• Let us consider some more examples to understand the linear
programming formulations better
Example 2
• A farmer is going to plant apples and bananas this year.
• It costs $ 40 per acre to plant apples and $ 60 per acre to plant bananas
• The farmer has a maximum of $ 7400 available for planting.
• To plant apples trees requires 20 labor hours per acre;
• To plant banana trees requires 25 labor hours.
• Suppose the farmer has a total of 3300 labor hours available.
• If he expects to make a profit of $ 150 per acre on apples and $ 200 per
acre on bananas,
• How many acres each of apples and bananas should he cultivate?
 Example 3
• A company makes two products (X and Y) using two machines (A and B). Each
unit of X that is produced requires 50 minutes processing time on machine A
and 30 minutes processing time on machine B. Each unit of Y that is produced
requires 24 minutes processing time on machine A and 33 minutes processing
time on machine B.
• At the start of the current week there are 30 units of X and 90 units of Y in
stock. Available processing time on machine A is forecast to be 40 hours and
on machine B is forecast to be 35 hours.
• The demand for X in the current week is forecast to be 75 units and for Y is
forecast to be 95 units. Company policy is to maximise the combined sum of
the units of X and the units of Y in stock at the end of the week.
• Formulate the problem of deciding how much of each product to make in the
current week as a linear program.
 Quiz 1
• Let us consider a company making a single product
• The estimated demand for the product for the next four months are 1000,
800, 1200, 900 respectively
• The company has a regular time capacity of 800 per month and an over time
capacity of 200 per month
• The cost of regular time production is $20 per unit and the cost of overtime
production is $25 per unit
• The company can carry inventory to the next month and the holding cost is $3
per unit per month
• The demand has to be met every month. Formulate a linear programming
problem for the above situation
• Step 1: Formulate the LP and construct a simplex tableau
• Add slack variables to represent unused resources, thus eliminating inequality
constraints
• Construct simplex tableau to evaluate various combinations of resources to determine
which mix will most improve your solution
• Use the slack variables in the starting basic variable mix
• Step 2: Find the sacrifice and improvement rows
• Values in the sacrifice row indicate what will be lost in per-unit profit by making a
change in the resource allocation mix
• Values in the improvement row indicate what will be gained per-unit profit by making a
change
• Step 3: Apply the entry criteria
• Find the entering variable and mark the top of its column with an arrow pointing
down
• The entering variable is defined as the current non-basic variable that will most
improve the objective function if its value is increased from 0
• If tie occurs arbitrarily choose one as the entering variable
• When no improvement can be found the optimal solution is represented by the
current tableau
• If no positive number appears entering the variable’s column, this indicates that one
or more constraints are unbounded
• Since it is impossible to have an unlimited supply of a resource, an unbounded
solution indicates that the LP problem was formulated incorrectly
• Step 4: Apply the exit criteria
• Using the current tableau’s exchange coefficient from the entering column, calculate
the following exchange ratio for each row as:
solution value/exchange coefficient
• The exchange ration tells you which variable is the limiting resource i.e., the resource
that would run out first
• Find the lowest nonzero and nonnegative value, this variable is the limiting resource
• The basic variable in this row becomes the exiting variable
• In case of identical alternatives choose arbitrarily one and mark the exiting variable
with an arrow pointing left
• Step 5: Construct a new simplex tableau
• Constructing a new tableau is a way to evaluate a new corner
• One variable will enter the basic mix and one variable will leave the basic mix and
become a non-basic variable.
• The operation of an entering variable and exiting variable is called Pivot operation
• The simplex method is made up of a sequence of such pivots and the pivot identifies
the next corner to be evaluated
• The new basic mix always differs from the previous basic mix by one variable (exiting
being replaced by the entering variable)
• To construct the new tableau, replace the exiting variable in the basic mix with the
new entering variable.
• Other basic variable remain unchanged
• Change the unit profit or unit loss column with the value of the new entering variable.
• Compute the new row values to obtain a new set of exchange coefficients applicable to
each basic variable
• Step 6: Repeat steps 2 through 5 until you no longer can improve the
solution
A basic solution has the following properties:
1. Each variable is designated as either a non-basic variable or a
basic variable.
2. The number of basic variables equals the number of functional
constraints (now equations). Therefore, the number of non-basic
variables equals the total number of variables minus the number
of functional constraints.
3. The nonbasic variables are set equal to zero.
4. The values of the basic variables are obtained as the
simultaneous solution of the system of equations (functional
constraints in augmented form). (The set of basic variables is often
referred to as the basis.)
5. If the basic variables satisfy the nonnegativity constraints, the
basic solution is a BF solution.