Q10: Solve following LP by using Simplex Method:

Minimize Z = 4x + y

Subject to,

3x + 4y ≥ 20

x + 5y ≥ 15

And x ≥ 0 y ≥ 0

Following instruction are not part of answer, they are just to explain you-

Just check vary carefully it is question of minimize and not of maximize, so very careful, do not overlook
in hurry.
Constraints sign are ≥ (Greater than or equal to) and not ≤ (Less than or equal to), so its handling is
different. Remember for ≤ (Less than or equal to), we add Slack variable to balance both side (Changing
inequalities into equalities) but for ≥ (Greater than or equal to):

Entire solution for this question is explained with lots of explanation (for theory question).

Answer:

Step 1: changing objective function from Minimization to Maximization by multiplying by -1:

- Z = -4x - y
0r

Max Z’= -Z = -4x -1y

Step 2: changing inequalities of constraints to equalities:

3x + 4y –S1 = 20 Left side is either more or equal to right side because

x + 5y –S2 = 15 3x + 4y should be at least 20,21,22………..
So to make both side equal this time we have to subtract
variable S1 and S2. So remember this rule that for ≥(Greater than or equal to), we will subtract one
variable from left side.’
S1 and S2 are known as known as “Surplus Variable” because left side has surplus (more) than Left
side.

Surplus Variables are also no-negative variables and do not contribute to objective function means
neither to cost and nor to profit. (Means if there is greater than or equal sign constraint in maximization
case)

Step 2: Revising entire LP:

Max Z’ = -4x - 1 y + 0 S1 + 0 S2

Subject to,

3x + 4y – 1S1 + 0 S2= 20

x + 5y+ 0 S1 – 1S2 = 15 and

x ≥ 0 y ≥ 0 S1 ≥ 0 and S2 ≥ 0
 Step 3: Finding Initial Basic Feasible Solution:

Putting x=0 and y =0 in revise constraints,

3 x 0 + 4 x 0 - 1S1 + 0= 20, so 1S1 = - 20 (minus 20)

x 0 + 5x 0+ 0 – 1S2 = 15, so 1S2 = - 15 (minus 15)

But both Surplus variables became negative, which is violation of non negative restriction as we know S1
≥ 0 and S2 ≥ 0.
So remove this problem, in minimization case there is one extra Step-4:

Step 4: Revising LP again and then Finding Initial Basic Feasible Solution:

To remove this problem, we have to introduce one another Artificial Slack Variable A1 and A2 for both
constraints.

Characteristics of Artificial Variable:

1. These are only imaginary variables and just introduce to remove violation of non negative
restriction of Surplus Variables. (arises in step 3)
2. These variables can not be present in final answer(Optimum Solution) in any case. As told you
in maximization case Slack Variable or in minimization case Surplus Variables can be remain in
Final Solution(Optimum Solution) when table preparation is stopped but Artificial variables
should be removed from Basic Solution Variable column and as soon it is removed, it can not
be re- entered like Slack Variable can be re-entered.
3. These Artificial variables is assigned an infinitely very large Cost denoted by M. As it is cost so
always remember:

a) For Minimization of Cost objective function, it will be added as it will increase cost
+ MA1 + MA2
b) While for Maximization of Profit objective function, it will be deleted as it will reduce
the profit as it is declared cost.
- MA1 - MA2

Re writing Revise LP after step 2:

Max Z’ = -4x - 1 y + 0 S1 + 0 S2- MA1 - MA2

Subject to,

3x + 4y – 1S1 + 0 S2+ 1A1 + 0A2 = 20

x + 5y+ 0 S1 – 1S2 + 0A1 + 1A2 = 15 and

x ≥ 0 y ≥ 0 S1 ≥ 0, S2 ≥ 0, A1 ≥ 0 and A2 ≥ 0

Now remember for ≥ (Greater than or equal to) sign, to find Initial feasible solution: Put Working
Variable x= 0, y =0 and also Surplus Variable S1 = 0, S2 = 0 in just revised LP:

3 x 0 + 4 x 0 - 0 + 0+ 1A1+ 0 = 20, so A1 = 20

x 0 + 5x 0+ 0 – 0+ 1A2 = 15, so A2 = 15.

 Step 5: Prepare Simplex Table-1 with Re written revise LP and A1 = 20 and A2 = 15 according
to revised LP:

Max Z’ = -4x - 1 y + 0 S1 + 0 S2- MA1 - MA2

Subject to,

3x + 4y – 1S1 + 0 S2+ 1A1 + 0A2 = 20

x + 5y+ 0 S1 – 1S2 + 0A1 + 1A2 = 15 and

x ≥ 0 y ≥ 0 S1 ≥ 0, S2 ≥ 0, A1 ≥ 0 and A2 ≥ 0

TABLE - 1

Contribution Per -4 -1 0 0 -M -M Minimum

Unit(Cj) Positive
Ratio
Contribution Basic Solution x y S1 S2 A1 A2
Per Unit of Solution Values
Solution Variables
Variables

-M A1 20 3 4 -1 0 1 0 20/4 = 5

-M A2 15 1 5 0 -1 0 1 15/5 = 3

Contribution Loss Per Unit (Zj) -4M -9M M M -M -M

Net Contribution Per Unit (Cj-Zj) -4 -1+9M 0-M 0-M 0 0

+4M
 No
te:
1. Remember for finding Key Column in Maximization case, it is Maximum value in (Cj-Zj).

So to find maximum value, some time difficult to find by just observation, so as we know M is
very very large quantiy, so let, it is M=100 (any value but should be large, 1000,1000 like that) and
now calculate value of Cj-Zj for each column and find Maximum Value:

For x Column = -4 +4M = -4 + (4 x 100) = 396

For y Column = -1+9M = -1 –+ (9x100) = -1 +900 = 899 (It is maximum value and be declared as Key
Column)

For S1 Column = -M= -100

For S2 Column = -M = -100

2. Remember for finding Key Row in both Maximization,it is Minimum Positive Ratio.

3. Table preparation will remain continue, till all the values of Cj-Zj will be become either 0
or Negative.

So just check still two column have positive values (-4 +4M and -1+9M), so table-II will be
prepared as we have done for Table-II.

TABLE - 2

Contribution Per -4 -1 0 0 -M Minimum

Unit(Cj) Positive
Ratio
Contribution Basic Soluti x y S1 S2 A1
Per Unit of Solution on
Solution Variables Value
Variables s

8/11/5=40/1
-M A1 8 11/5 0 -1 4/5 1
1

5/5 3/1/5=
-1 y 3 1/5 0 -1/5 0
=1 15

Contribution Loss Per Unit (Zj) -11M- -1 M - -M

1/5 4M+1/5

Net Contribution Per Unit (Cj-Zj) -4 +11 M 0 -M 4M-1/5 0

+1/5
 We have found:

1. Key Column and Substitute Variable: It is (- 4 x)

2. Key Row and Departing Variable: It is (- M A1)

3. Key Element : Intersection of Key Column and Key Row : 11/5

4. Non Key Row: (-1 y)

5. Corresponding value for Non Key Row: 4

Now try to make Table-3

TABLE - 3

Contribution Per -4 -1 0 0 Minimum

Unit(Cj) Positive
Ratio
Contribution Basic Soluti x y S1 S2
Per Unit of Solution on
Solution Variables Value
Variables s

-4 x

-1 y

Contribution Loss Per Unit (Zj)

Net Contribution Per Unit (Cj-Zj)