▪ Linear programming

▪ Different forms of linear programming

- General form of linear programming
- Standard / canonical form of linear programming
- Matrix standard form of linear programming
- Solution form of linear programming

▪ Characteristics of standard form of linear programming

▪ Transformation to a standard form of linear programming
▪ Assumptions of linear programming
- Proportionality
- Additivity
- Divisibility
- Certainty

▪ Usefulness of linear programming

▪ Limitations of linear programming
▪ Linear programming formulation
▪ Solution of linear programming
- Graphical method (feasible region, feasible solution, optimum solution, most favorable
value)

- Simplex method (with big M method)

- Sensitivity analysis
- Duality

Dr. Md. Ashik-E-Rabbani, Professor, Department of Farm Power and Machinery, BAU
Linear programming:
Linear programming involves the planning of activities to obtain an
optimum result, i.e a result that reaches the specified goal best (according to the
mathematical model) among all feasible alternatives.

Different forms of linear programming

A. General form of linear programming
B. Standard / canonical form of linear programming
C. Matrix standard form of linear programming
D. Solution form of linear programming
A. General form of linear programming:
Max (or min), Z = c1 x1 + c2 x2 + ----------------+ cn xn
Subject to,
a11 x1+ a12 x2 + ----------------+ a1 n xn ≤ b1
a21 x1+ a22 x2 + ----------------+ a2 n xn ≤ b2
……………………………………
am1 x1+ am2 x2 + ----------------+ amn xn ≤ bm
x1, x2, ---------------- xn ≥ 0

B. Standard / canonical form of linear programming:

Max, Z = j

Subject to,
Where,
aij xj ≤ bi i = 1, 2, 3,…………m
j = 1, 2, 3,………….n
xj ≥ 0
Where,
i = refers to ith resources
j = jth activity
bi = available amount of resource, i
aij = amount of resource i that must be denoted to activity, j

Dr. Md. Ashik-E-Rabbani, Professor, Department of Farm Power and Machinery, BAU
cj = worth/ cost per unit activity, j
xj = decision variable
m = number of constraints
n = number of decision variables
Characteristics of standard form of linear programming:
Three characteristics of standard form
1. All decision variables are non-negative;
2. All constraints are of the less than or equal to type and
3. The objective function is maximization.

Transformation to a standard form of linear programming:

There are three types transformation to standard form:-
1. Changing minimization to maximization
Minimizing, Y = c1 x1 + c2 x2 +
is the same as
Maximizing, Z = - Y = - c1 x1 - c2 x2 -
2. Changing the direction of an equality
a11 x1+ a12 x2 + ----------------+ a1 n xn ≥ b1
is the same as
- a11 x1- a12 x2 - ---------------- - a1 n xn ≤ b1
3. Equality constraint
a11 x1+ a12 x2 + ----------------+ a1 n xn = b1
is equivalent to two constraints
a11 x1+ a12 x2 + ----------------+ a1 n xn ≤ b1
- a11 x1- a12 x2 - ---------------- - a1 n xn ≤ b1

Dr. Md. Ashik-E-Rabbani, Professor, Department of Farm Power and Machinery, BAU
 C. Matrix standard form of linear programming:
Max Z = C X
Subject to, constrains A X ≤ b
X ≥0

Where, X =
[] b=
[]
A=
[ ]
C = [ C1 C2 . . . Cn ]

D. Solution form of linear programming:

P - 40X1 - 30X2 - 0 X3 - 0 X4 - 0 X5 = 0
Subject to,
X1 +X3 =16
X2 +X4 =8
X1 + 2X2 +X5 = 24
X1, X2, X3, X4, X5 ≥ 0

Assumptions of Linear programming

(1) Proportionality assumption:
Variables in LP models are assumed to exhibit proportionality. Proportionality deals
with the contribution per unit of each decision variable to the objective function. This
contribution is assumed constant and independent of the variable level. Similarly, the
use of each resource per unit of each decision variable is assumed constant and
independent of variable level. There are no economies of scale.

Dr. Md. Ashik-E-Rabbani, Professor, Department of Farm Power and Machinery, BAU
For example, in the general LP problem, the net return per unit of Xj produced is cj. If
the solution uses one unit of Xj, then cj units of return are earned, and if 100 units are
produced, then returns are 100cj. Under this assumption, the total contribution of Xj
to the objective function is always proportional to its level.

(2) Additivity assumption:

Additivity deals with the relationships among the decision variables. Simply put their
contributions to an equation must be additive. The total value of the objective
function equals the sum of the contributions of each variable to the objective
function. Similarly, total resource use is the sum of the resource use of each variable.
This requirement rules out the possibility that interaction or multiplicative terms
appear in the objective function or the constraints.
In the general LP formulation, when considering variables X j and Xk, the value of the
objective function must always equal cj times Xj plus ck times Xk. Using Xj does not
affect the per unit net return of Xk and vice versa. Similarly, total resource use of
resource I is the sum of aij Xj and aik Xk. Using Xj does not alter the resource
requirement of Xk.

(3) Divisibility assumption:

The problem formulation assumes that all decision variables can take on any non-
negative value including fractional ones; (i.e., the decision variables are continuous).

(4) Certainty assumption:

The certainty assumption requires that the parameters cj, bi, and aij be known
constants. The optimum solution derived is predicated on perfect knowledge of all
the parameter values. Since all exogenous factors are assumed to be known and
fixed, LP models are sometimes called non- stochastic as contrasted with models
explicitly dealing with stochastic factors. This assumption gives rise to the term
"deterministic" analysis.

Usefulness of Linear programming

Linear programming techniques are widely used to solve of various problem i.e,
military, economic, industrial and social problem etc. Three primary reasons for its
wide use-

Dr. Md. Ashik-E-Rabbani, Professor, Department of Farm Power and Machinery, BAU
 ▪ A large variety of problems in diverse field be represented or at least
approximated as linear programming model.
▪ Efficient techniques for solving linear programming problems are available.
▪ Ease through which data variation (sensitivity analysis) can be handled
through linear programming model.

Limitations of Linear programming

The limitations of the linear programming model are-
▪ All the mathematical function in the linear programming model must be
linear function.
▪ Linear programming model assumes that all the parameters of the model
are known constant. But real situation, this assumption is seldom satisfied
precisely.
▪ Linear programming model is the only mathematical model aim to
idealize the real situation.

Dr. Md. Ashik-E-Rabbani, Professor, Department of Farm Power and Machinery, BAU
 Liner programming formulation
1. A Farm grows Jute and Wheat on its 500 acre of land. An acre of Jute brings Tk.
100 profit and an acre of Wheat brings Tk. 200 profit. Because of Govt. regulation
no more than 200 acres can be planted in Jute. During the planting season 1200
man-hrs of planting time will be available. Each acre of Jute requires 2 man-hrs
while each acre of Wheat requires 6 man-hrs. How many acres of Jute and how
many acres of Wheat should be planted in order to maximum profit.

Solution:
Let,
X1 = amount of land for Jute, acre
X2 = amount of land for Wheat, acre
Objective function,
Maximize profit P = 100X1 + 200X2
Subject to,
X1 + X2 ≤ 500
X1 ≤ 200
2 X1 + 6 X2 ≤ 1200
X1, X2 ≥ 0
2. A manufacturing firm has three departments, namely cutting, mixing and packing
departments to produce two products, A and B. The product A and B are sold to
the market with net profit of Tk. 40 per ton and Tk. 30 per ton respectively. Each
ton of product A requires 30 minutes for cutting, 20 minutes for packing. Each ton
of product B requires 60 minutes for mixing, 40 minutes for packing. The daily
capacities of the manufacturing plant are 8 hours. Determine the optimum
quantity of A and B to be produced per day to maximize profit.
Solution:
Let,
X1 = amount of product A to be produced per day, ton
X2 = amount of product B to be produced per day, ton
Objective function,
Maximize profit P = 40X1 + 30X2
Subject to,
½ X1 + 0 X2 ≤ 8
0 X1 + 1 X2 ≤ 8

Dr. Md. Ashik-E-Rabbani, Professor, Department of Farm Power and Machinery, BAU
 ⅓ X1 + ⅔X2 ≤ 8
X1, X2 ≥ 0
3. Two foods A and B contain Calcium, protein and Calories. Each kg of food A
contains 10 units of calcium, 5 units of protein and 2 units of calories. Each kg of
food B contains 4 units of calcium, 5 units of protein and 6 units of calories. The
minimum daily requirement of calcium, protein and calories are 20 units, 20 units
and 12 units respectively. The cost of food A and B if Tk. 0.6 /kg and Tk. 1 /kg.
Determine the optimum amount of each food combination, which will satisfy the
specified nutritional requirements and minimize the total cost of the diet.
Solution:
Let,
X1 = amount of food A to be consumed per day, kg
X2 = amount of food B to be consumed per day, kg
Objective function,
Minimum cost C = 0.6X1 + 1.0X2
Subject to,
10 X1 + 4 X2 ≥ 20
5 X1 + 5 X2 ≥20
2 X1 + 6 X2 ≥ 12
X1, X2 ≥ 0
4. The management of Padma Oil Company has received orders from Bangladesh
Railway to supply 600 units of diesel and 500 units gasoline. Padma Oil Company
has two processes that can be used (process A and process B ) for manufacturing
of diesel and gasoline. One hour of process A produces 5 units of diesel and 10
units of gasoline. One hour of process B produces 8 units of diesel and 6 units of
gasoline. Process A uses a particular combination of crude’s which cost Tk. 7 per
hour, Process B uses a different combination of crude’s which cost Tk. 9.20 per
hour. Find the optimal mix of process A and Process B.
Solution:
Let,
X1 = No. of hrs process A to be run
X2 = No. of hrs process B to be run
Objective function,
Minimum cost C = 7X1 + 9.2X2
Subject to,
5 X1 + 8 X2 = 600

Dr. Md. Ashik-E-Rabbani, Professor, Department of Farm Power and Machinery, BAU
 10 X1 + 6 X2 = 500
X1 , X2 ≥ 0
Hints:
Availability / Daily capacity / More than use ≤
Minimum / must produced / Minimum daily requirement use ≥
Supply / The company has use =

Liner programming graphical solution

The graphical method follows these steps:
1. Change inequalities to equalities.
2. Graph the equalities.
3. Identify the correct side for the original inequalities.
4. Then identify the feasible region, the area of Feasible Solution.
5. Determine the profit or cost at each of the corner points (basic feasible
solutions) of the feasible region.
6. Pick either the most profitable or least cost combination, which is an Optimal
Solution.
Feasible region: The feasible region is the collection of all feasible solution.
Feasible solution: A feasible solution is a solution for which all constraints are
satisfied.
Optimum solution: An optimum solution is a feasible solution that has the most
favorable value of the objective function.
Most favorable value: Most favorable value means the largest and smallest value
depending upon whether the objective function is maximization or minimization.
For example,
Maximize profit CM = $25A + $40B
Subject to,
2A + 4B ≤ 100
3A + 2B ≤ 90
A ≥ 0, B ≥ 0
After going through steps 1 through 4, the feasible region (shaded area) is obtained,
as shown in the following exhibit. Then all the corner points in the feasible region are
evaluated in terms of their CM as follows:

Dr. Md. Ashik-E-Rabbani, Professor, Department of Farm Power and Machinery, BAU
Dr. Md. Ashik-E-Rabbani, Professor, Department of Farm Power and Machinery, BAU
 Liner programming Simplex method solution

1. Maximize profit P = 40X1 + 30X2

Subject to,
X1 ≤ 16
X2 ≤ 8
X1 +2X2 ≤ 24
X1, X2 ≥ 0
Find the optimum value of X1, X2 for maximization of profit P?
Solution:
P - 40X1 - 30X2 - 0 Y1- 0 Y2- 0 Y3 = 0 ---------------------------------(1)
Subject to,
X1 +Y1 =16 -------------(2)
X2 +Y2 = 8 --------------(3)
X1 + 2X2 +Y3 = 24 -------------(4)
Y1, Y2, Y3 are slack variables
X1, X2, Y1, Y2, Y3 ≥ 0
EV

B.V E.N P X1 X2 Y1 Y2 Y3 R.H.S Ratio

P 1 1 -40 -30 0 0 0 0 -
Y1 2 0 1 PE 0 1 0 0 16 16 (mini.)
Y2 3 0 0 1 0 1 0 8 -
Y3 4 0 1 2 0 0 1 24 24
EV
P 1 1 0 -30 40 0 0 640 -
X1 2 0 1 0 1 0 0 16 -
Y2 3 0 0 1 0 1 0 8 8
Y3 4 0 0 2 PE -1 0 1 8 4(mini.)

P 1 1 0 0 25 0 15 760
X1 2 0 1 0 1 0 0 16
Y2 3 0 0 0 1/2 1 - 1/2 4
X2 4 0 0 1 - 1/2 0 1/2 4
There are no negative values in step 3 so this is the optimum solution,
P = 760, X1 = 16, X2 = 4, Y1 = 0, Y2 = 4 and Y3 = 0

Dr. Md. Ashik-E-Rabbani, Professor, Department of Farm Power and Machinery, BAU
Sensitivity Analysis:
In general means, the relative magnitude of change in the measure of merit
caused by one or more changes in estimated elements or variables.

If the resource of constraint no. 1 be changed from 16 to 17 what would be the new
value of P?
Answer:
In the step 3,
The equation number 1 shows the value of Y1 is 25. Therefore,
Profit, P = 760+25 =785

Duality:

Primal Solution Dual Solution

Maximize profit P = 40X1 + 30X2 Minimize cost, C = 16Y1 +8Y2 +24Y3

Subject to, Subject to,
X1 ≤ 16 Y1 +Y3 ≥ 40
X2 ≤ 8 Y2 +2Y3 ≥ 16
X1 +2X2 ≤ 24 Y1, Y2 ≥ 0
X1, X2 ≥ 0

Result from Primal to Dual

P = 760, X1 = 16, X2 = 4, Y1 = 0, Y2 = In the step 3,

4 and Y3 = 0 The equation number 1 shows the
value of
C = 760, Y1 = 25, Y2 = 0, Y3 = 15, X1
= 0 and X2 = 0

Dr. Md. Ashik-E-Rabbani, Professor, Department of Farm Power and Machinery, BAU
 2. Maximize profit Z = 4X1 + 3X2 + 7X3
Subject to,
2X1 + X2 + 3X3 ≤ 120
X1 + 3X2 + 2X3 = 120
X1, X2, X3 ≥ 0
Find the optimum value of X1, X2, and X3 for maximization of profit Z?
Solution:
Z - 4X1 -3X2 -7X3 - 0X4 + MX5 = 0----------(1)
Subject to,
2 X1 + X2 + 3X3 +X4 =120--------(2)
X1 + 3X2 +2X3 +X5 =120 --------(3)
X1, X2, X3, X4 ≥ 0; X5 = 0
B.V E.N Z X1 X2 X3 X4 X5 R.H.S Ratio
1 1 -4 -3 -7 0 M 0
Unadjusted 2 0 2 1 3 1 0 120
3 0 1 3 2 0 1 120
EV
Z 1 1 -4-M -3-3M -7-2M 0 0 -120M -
X4 2 0 2 1 3 1 0 120 120
X5 3 0 1 3 PE 2 0 1 120 40 (mini.)
EV
Z 1 1 -3 0 -5 0 1+M 120 -
X4 2 0 5/3 0 7/3 PE 1 -1/3 80 240/7(mini.)
X2 3 0 1/3 1 2/3 0 1/3 40 60

Z 1 1 4/7 0 0 15/7 2/7+M 2040/7

X3 2 0 5/7 0 1 3/7 -1/7 240/7
X2 3 0 -2/3 1 0 -2/7 3/7 120/7

There are no negative values in step 4 so this is the optimum solution,

Z = 2040/7, X1 = 0, X2 = 120/7, X3 = 240/7, X4 = 0 and X5 = 0

Dr. Md. Ashik-E-Rabbani, Professor, Department of Farm Power and Machinery, BAU
 3. Maximize profit Z = 3X1 + 5X2
Subject to,
X1 ≤ 4
X2 ≤ 6
3X1 +2X2 ≥ 18
X1, X2 ≥ 0
Find the optimum value of X1, X2 for maximization of profit Z?
Solution:
Z - 3X1 - 5X2 - 0 X3- 0 X4- 0 X5 +MX6 = 0 --------------(1)
Subject to,
X1 +X3 = 4 --------------(2)
X2 +X4 = 6 --------------(3)
3X1 + 2X2 -X5 +X6 = 24 ------------(4)
X1, X2, X3, X4, X5 ≥ 0; X6 = 0

B.V E.N Z X1 X2 X3 X4 X5 X6 R.H.S Ratio

1 1 -3 -5 0 0 0 M 0
2 0 1 0 1 0 0 0 4
Unadjusted
3 0 0 1 0 1 0 0 6
4 0 3 2 0 0 -1 1 18
EV
Z 1 1 -3-3M -5-2M 0 0 M 0 -18M -
X3 2 0 1 PE 0 1 0 0 0 4 4 (mini.)
X4 3 0 0 1 0 1 0 0 6 -
X6 4 0 3 2 0 0 -1 1 18 6
EV
Z 1 1 0 -5-2M 3+3M 0 M 0 12-6M -
X1 2 0 1 0 1 0 0 0 4 -
X4 3 0 0 1 0 1 0 0 6 6
X6 4 0 0 2 PE -3 0 -1 1 6 3 (mini.)
EV
Z 1 1 0 0 -4.5 0 -2.5 2.5+M 27 -
X1 2 0 1 0 1 0 0 0 4 4
X4 3 0 0 0 3/2 PE 1 1/2 -1/2 3 2 (mini.)
X2 4 0 0 1 -3/2 0 -1/2 1/2 3 -
EV
Z 1 1 0 0 0 3 -1 M+1 36 -
X1 2 0 1 0 0 -2/3 -1/3 1/3 2 -
X3 3 0 0 0 1 2/3 1/3 PE -1/3 2 6 (mini.)
X2 4 0 0 1 0 1 0 0 6 -

Z 1 1 0 0 3 5 0 M 42
X1 2 0 1 0 1 0 0 0 4
X5 3 0 0 0 3 2 1 -1 6
X2 4 0 0 1 0 1 0 0 6

There are no negative values in step 6 so this is the optimum solution,

Z = 42, X1 = 4, X2 = 6, X3 = 0, X4 = 0, X5 = 6 and X6 = 0

Dr. Md. Ashik-E-Rabbani, Professor, Department of Farm Power and Machinery, BAU
 Assignment
1. Graphical Solution:
Maximize profit P = 40X1 + 30X2 Minimize cost, C = 0.6Y1 +Y2
Subject to, Subject to,
X1 ≤ 16 10Y1 +4Y2 ≥ 20
X2 ≤ 8 5Y1+5Y2 ≥ 20
X1 +2X2 ≤ 24 2Y1 +6Y2 ≥ 12
X1, X2 ≥ 0 Y1, Y2 ≥ 0

2. Simplex Method:
a. Maximize profit P = 2X1 + 5X2 b. Maximize profit P = 100X1 + 200X2
Subject to,
Subject to,
X1 + X2 ≤ 500
X1 ≤ 4
X1 ≤ 200
X2 ≤ 3
2 X1 + 6 X2 ≤ 1200
X1 +2X2 ≤ 8 X1, X2 ≥ 0
X1, X2 ≥ 0

c. Maximize profit P = 5X1 + 3X2 +4X3 d. Minimize cost, C = 0.6Y1 +Y2

Subject to,
Subject to,
2X1 +5X2 ≤6
10Y1 +4Y2 ≥ 20
2X1 +3X2+ X3 ≥ 5
5Y1+5Y2 ≥ 20
3X1 +4X3 =3
X1, X2, X3 ≥ 0 2Y1 +6Y2 ≥ 12
Y1, Y2 ≥ 0

Dr. Md. Ashik-E-Rabbani, Professor, Department of Farm Power and Machinery, BAU
1. A food processing plant manufactures hotdog and burger. They grind their own flour for the
burger at a maximum rate of 200 kg per week. Each burger requires 0.1 kg of flour. They
currently have a contract with Macdonald Co. which specifies that a delivery of 800 kg of beef
is delivered every Monday. Each hotdog requires 0.25 kg of beef. Five employees working full
time (40 hours per week). Each hotdog requires 3 minutes of labour and each burger requires 2
minutes of labour. Each hotdog yields a profit of Tk. 0.20 and each burger yields a profit of Tk.
0.10. The company wants to know how many hotdogs and burger they should produce each
week to achieve the highest possible profit.
2. A manufacturer produces three products A, B and C. Each product can be produced on either of
the two machines I and II. The time required to produce one unit of each product on a machine
is given in the table.
Time to produce one unit (hrs)
Product
Machine I Machine II
A 0.5 0.6
B 0.7 0.8
C 0.9 1.05
There are 850 hours available on each machine. The operating cost is Tk. 5/hr for Machine I
and Tk. 4/hr for Machine II. The market requirements are at least 90 units of A, at least 80 units
of B and at least 60 units of C. The manufacturer wishes to meet the requirement at minimum
cost. Formulate this problem as LP problem.
3. A farmer has a 100 acre farm. He can sell the tomatoes, lettuces and radishes, he can rise. The
prices he can obtain at Tk. 1 per kg of tomatoes, Tk. 0.75 per head of lettuces and Tk. 2 per kg
of radishes. The average yield per acre is 2000 kg of tomatoes, 3000 heads of lettuce and 1000
kg of radishes. Fertilizer is available 1000 kg at price of Tk. 0.50 per kg for this season’s. The
amounts required per acre are 100 kg each for tomatoes and lettuce and 50 kg for radishes.
Labour required for sowing, cultivating and harvesting per acre are 5 man-days for tomatoes
and radishes and 6 man-days for lettuce. A total of 400 man-days of labour available at Tk. 20
per man-day. Formulate an LP model for this problem in order to maximize the farmer’s total
profit.

Dr. Md. Ashik-E-Rabbani, Professor, Department of Farm Power and Machinery, BAU