Higher Engineering Mathematics

Prof. P. N. Agrawal
Department of Mathematics
Indian Institute of Technology, Roorkee
Lecture 46 – Big M Method - I

Hello friends! Welcome to my lecture on Big M Method, first lecture on Big M Method. This
method was given by A. Charnes and it consists of the following steps:

First we express the problem in the standard form. So by expressing the problem in the
standard form means we first put the problem in the maximization form, and then the second
thing is that we ensure that all the bi , s in the constraints are non-negative. Third thing that we
have to do is that we have to express all the constraints in the form of equations by using
slack or surplus variables. The fourth thing is that we have to see that all the variables in the
problem are non-negative.

(Refer Slide Time: 1:28)

So first we put the problem in the standard form. Then the second step we do is that we add
non-negative variables to the left hand side of all those constraints which are of ≥ type. Such
new variables are called artificial variables and the purpose of introducing these variables is
just to obtain an initial basic feasible solution.

So these variables are causing violation of the corresponding constraints, so therefore we

would like to get rid of these variables and would not allow them to appear in the final
solution. And that is why we assign a very large penalty, - M to these artificial variables in
the objective function. So purpose of introducing objective….artificial variables is just to
obtain a initial basic feasible solution. But they violate the corresponding constraints,
therefore we would like to get rid of them and also would not allow them to appear in the
final solution.

(Refer Slide Time: 2:27)

Now what we do is let us solve the modified LPP by using simplex method. At any iteration
of the simplex method, the following three cases may arise: First case is that, there remains
no artificial variable in the basis and the optimality condition is satisfied. Then the solution is
an optimal basic feasible solution to the problem.

The other case could be, there is at least one artificial variable in the basis at zero level, that
means zero value in the b-column. The artificial variable having zero value in the b-column
and the optimality condition is satisfied, then we say that the solution is a degenerate optimal
basic feasible solution.
(Refer Slide Time: 3:13)

Now the third case is there is at least one artificial variable in the basis at non-zero level. That
means with the positive value in the b-column and the optimality condition is satisfied, then
the problem has no feasible solution because the artificial variable, there is at least one
artificial variable in the basis at non-zero level and so the final solution is not optimal because
the objective function contains an unknown quantity M.

Such a solution satisfies the constraints but does not optimize the objective function and
therefore is called pseudo optimal solution. Now in the step 4, we continue the simplex
method until either an optimal basic feasible solution is obtained or an unbounded solution is
indicated.

(Refer Slide Time: 4:07)

Now let us consider this example and solve this example by using Charne’s method. So use
Charne’s penalty method to minimize Z = 2 x 1 + x2 , subject to 3 x 1+ x2 = 3; 4 x 1+ 3 x2 ≥ 6; x 1+
2 x2 ≤ 3; x 1, x2 ≥ 0. Since this equation, this LPP is of minimization type, first we convert it to
maximization type and therefore consider, we consider maximum of Z ' .

(Refer Slide Time: 4:42)

Z is - Z. So Z = 2 x 1+ x2 , so this is – 2x 1- x2 . And then we assign 0 values to this, because the

'

first constraint is of the type of equality. 3 x 1 + x2 = 3, this is first constraint. Second is 4 x 1 +

3 x2 ≥ 6; x 1 + 2 x2 ≤ 3. Now this is of equality type, so we have to add an artificial variable
here. And here this is of ≥ type, so we add one, we subtract surplus variable as s1 let us say
and add an artificial variable here. And here we add a slack variable.
So let us say we have, this is your surplus variable s1, this is your slack variable s2. And A 1
and A 2 are artificial variables which have been given the value - 1, that is the penalty - M. So
the equations, first equation, 3 x 1 + x2 = 3 can then be written as 3 x 1+ x2 + 0 s1, 0 s2 + A 1 + 0 A 2
= 3. We are adding artificial variable here. Now second constant is 4 x 1+ 3 x2 ≥ 6, so we
consider, we take a surplus variable and we subtract the surplus variable here.

So 4 x 1+ 3 x2 - s1+ 0 s2+ 0 A 1+ A 2. Because it is ≥ type, so we also have to take artificial

variable which we can take as A 2, A 2= 6. And then third constant is ≤ type, so x 1 + 2 x2 ≤ 3,
so x1 +¿ 2 x2 +0 s1 + s2 + 0 A 1 + 0 A 2 = 3. So s2 is the slack variable here. And we take x 1, x2 ,
s1, s2, A 1, A 2 to be all non-negative.

So thus we consider the corresponding maximization problem where we have to maximize Z'
= - 2 x 1 - x2 + 0 s1 + 0 s2 - M A 1 - M A 2 , subject to these equations, these three equations,
where the variables x 1, x2 , s1, s2, A 1, A 2are all ≥ 0. So the second and third inequalities are
converted into equations by introducing surplus variable and slack variable. And first and
second constraints, this one and this one, they were of equality and ≥ type, so we have
introduced two artificial variables, A 1and A 2.
(Refer Slide Time: 7:58)

Now let us go to finding initial basic feasible solution. So surplus, now let us note that
surplus variable s1 is not a basic variable here.

(Refer Slide Time: 8:07)

This surplus variable s1 is not a basic variable here because if we take x 1, x2 , s2, A 1, A 2= 0, if
we take x 1=0, x2 = 0, and s2 = 0, A 1 = 0, then A 2 = 0, then what we get? s1 = - 6. So taking x 1
= 0, x2 = 0, s2 = 0, we have three equations here and there are how many constraints? 1, 2, 3,
4, 5, 6 constraints. So we need to take three constraints =0. So we can take x 1, x2 =0 and A 2
=0.
See, we have three equations here, 1, 2, 3. So to obtain initial basic feasible solution we need
to take three constraints =0. So that means let us take x 1 =0, x2 =0, A 2 =0. If we take x 1 =0, x2
=0, and A 2 =0, then we arrive at s1 =- 6. But s1 is ≥ 0, so what we have?

(Refer Slide Time: 9:31)

s 1 will not be a basic variable because its value is - 6. So as negative quantities are not
feasible, s1 must be prevented from appearing in the initial solution. For a feasible solution all
the variables must be non-negative. So here s1 must be prevented from appearing in the initial
solution and this we can do by taking s1 = 0. So by setting the other non-basic variables x 1, x2
each to = 0, so what we do now?

We take x 1 = 0, x2 = 0, and s1 = 0. x 1 = 0, x2 = 0, and s1 = 0, we obtain the initial basic feasible

solution as x 1 = 0, x2 = 0, s1 = 0, and then we get A 1 = 3.
(Refer Slide Time: 10:26)

A 1 equal to…here you see, x 1 is 0, x2 is 0 and then we are takings1 = 0. So we get A 1 = 3, we

get this equation. This equation gives you A 1 = 3. And here we get x 1, 0; x2 , 0; s1, 0, so A 2 =
6. And we get x 1, 0; x2 , 0 and yeah, so x 1, 0; x2 , 0 and this what we get? Here we have s2 =3.
x 1 + 2 x2 + 0s1 + s2 + 0 A 1 + 0 A 2 =3 gives s2 =3.

(Refer Slide Time: 11:15)

So what we have is this: The initial basic feasible solution is x 1 =0, x2 =0, s1 =0, A 1 =3, A 2 =6,
s2 =3.
(Refer Slide Time: 11:26)

And we then have the initial simplex table as: So these are C js, C js are the coefficients of
the variables in the objective function. So x 1, coefficient of x 1 is - 2, coefficient of x2 is - 1,
coefficient of s1 is 0, coefficient of s2 is 0, coefficient of A 1 is - M, coefficient of A 2 is - M.
These are the coefficients of x 1, x2 , s1, s2, A 1, A 2 in the objective function. The basis is A 1, A 2,
s2. Basis is A 1, A 2, s2.

(Refer Slide Time: 11:59)

(Refer Slide Time: 12:07)

This one, basic variables are A 1, A 2 and s2. And non-basic variables are x 1, x2 , s1. So non-
basic variables are x 1, x2 , s1. Okay, basic variables are s2, A 1, A 2, and these are the
coefficients which we have written from the constraints, the equations.

(Refer Slide Time: 12:20)

These equations. This equation, this equation and this equation: first equation, second
equation, third equation.
(Refer Slide Time: 12:31)

So the coefficients of x 1, x2 , s1, s2, A 1, A 2 are written here: 3, 1, 0, 0, 1, 0, 4, 3, - 1, 0, 0, 1, 1,

2, 0, 1, 0, 0. And these are the…This is the b column. b column is your 3, 6, 3, that is b
column. Now the coefficient of A 1 in the objective function is - M. So in the CB column we
write the coefficients of the basic variables, so - M, the coefficient of A 2 is - M and
coefficient of s2 is 0.

Now we calculate Z j. Z j =Sigma CB a i j. And so then let us see. This is your CB column.
So - M into 3, so we get - 3 M. - M into 4, we get - 4 M. And then 0 into 1 is 0. So we get
Sigma CB a i j is =- 7 M. Then Sigma CB a i j we find for the second column. So - M into 1,
- M. - M into 3, - 3 M. And 0 into 2 is 0. So we get total - 4 M. So this is - 4 M.

And then we write similarly. - M into 0 is 0. - M into 1 is - M. This is 0 into 0, so we get….-

M into - 1 is M. So we get M here. Similarly this column s2, - M into 0 is 0. - M into 0 is 0. 0
into 1 is 0. So total sum is =0. And then - M into 1 is 0, - M into 1 is - M. - M into 0 is 0. 0
into 0 is 0. So sum is - M. So we get here - M.

And then - M into 0, - M into 1, 0 into 0; so we get sum as - M. And then here when you
multiply - M, 0 to the b column we get - 3 M, - 6 M and then 0. So total is - 9 M. Now we
find C j. Capital C j =small c j - Z j. So this is your C j. And this is here, this is our Z j. So - 2
- of - 7 M gives you 7 M - 2. So - 2 - - of 7 M gives us 7 M - 2.

Similarly - 1, - of - 4 M gives us 4 M - 1. And 0 - M, 0 - M is - M, so we get - M here. Then 0

- 0 is 0. Then - M - of - M is 0, so we get 0 here. Then - M - of - M is 0, so we get 0 here.
Now let us see. So C j is =c j - Z j. And here it is 7 M - 2, here it is 4 M - 1. Here it is - M. M
is a very large quantity, so we see that C j is positive. C j is positive in this column
corresponding to x 1 and column corresponding to x2 .

But since M is large, 7 M - 2 is more than 4 M - 1. So maximum C j is 7 M - 2. C j is positive

in the two columns, in the columns corresponding to x 1 and x2 . But since M is large 7 M - 2
is more than 4 M - 1. So 7 M - 2 is our key column, this is our key column. And then what we
have? We have to find the key row. So the elements of key column we divide to the elements
of b, the b column. The elements of b column are divided by the corresponding elements of
key column.

So 3 over 3 gives you 1 and then 6 over 4 gives you 3 by 2. So this is 1, this is 3 by 2 and
then 3 divided by 1 is =3. Now we have to find minimum of these three values. So minimum
is clearly 1. And therefore this row, this row is key row. This row is a key row. This arrow
shows the key row. And at the intersection of key column and key row we get the key
element, so this is our key element.

(Refer Slide Time: 17:31)

(Refer Slide Time: 17:43)

So we have the following: Since C j is positive under x 1, x2 column, this is not an optimal
solution. We therefore iterate towards finding optimal solution. And what we do then? We
have found the key element. Now key element, we divide by the key element all the elements
of the key row. So to get here 1, so we divide this key row by 3. We get 1 here, 1 by 3 here, 0
here, 0 here, 1 by 3 here and 1 here. So we get that.

1, 1 by 3, 0, 0, 0, 1. So we divide the elements of the key row by the key element to obtain
this key element as 1. And then we use elementary row operations to make all the other
entries in the key column =0. So by using….now this is what? This 4. So we multiply this
row by 4, key row by 4 and subtract from the second row to have 0 here. And then to get 0
here, we subtract the key row, this key row, this row from the third row. So what we do?

After dividing this key row by 3, we make the element 3 as 1 here and then with the help of
this 1 we make the elements in the key column, that is 4 and 1, 0, using elementary row
operations. To make 4, 0 we have to multiply the elements of this key row by 4 and subtract
from the second row. And then to make this 1, 0 we subtract the elements of the key row by, I
mean elements of the key row from the elements of this third row after the key element has
been made 1.

So that we have to do here. So key row after dividing by 3 becomes 1, 1 by 3, 0, 0, 0, 1. And

then we subtract four times of this from the second row. We get 0, 5 by 3; how we get 5 by 3?
We can see. See, this is 1 by 3. After dividing by 3 this becomes 1 by 3. And then we
multiply it by 4 and subtract from this. So we will get what? 3 - 4 by 3. So that means 9 - 4 by
3. So 5 by 3. So this element, this will become 0, this will become 5 by 3.

Here we will get - 1, - 0 into we are dividing by 3, so 0 by 3 is 0. And then multiplying by 4,

so 0 into 4, so we get - 1. So this element will remain - 1. This element here will also remain
0 because we are dividing this row by 3 and then multiplying by 4 and subtracting from this
row. But here what will happen? This is 0. We are dividing it by 3, we are dividing it by 3
and then multiplying by 4 and subtracting, so 0 - 4 by 3.

So this will be - 4 by 3. So this A 1 element here will be - 4 by 3. So what we do is…now

what will happen is that this is incoming variable, x 1 is incoming variable and A 1 is outgoing
variable. So in the new simplex table A 1 will be replaced by x 1 and the coefficient of x 1 that
is - 2 will come in place of - M.

So here we get - 2 and then A 1 is replaced by x 1. A 1 is replaced by x 1 and the column

corresponding to A 1 is deleted. There is no column in the new simplex table corresponding to
A 1. A 1 is the artificial variable. We want to get rid of A 1 as quickly as possible. So after
writing this incoming variable x 1 and the corresponding value in the CB column, we make
this element 1.

We make this element…key row…key element 1 and then subtract it from the….using
elementary row operations from the second and third rows to make this 0 and this 0. And
what we get is this: Yeah, so we have….column corresponding to A 1 is not there. A 2, A 2 we
have to look into, 1 - 0 by 3 into 4, 4 into 0 by 3, so we get 1 here. So we get 1 here. And then
b column, in the b column what we will get? In the b column we have 6 -, we divided it by 3
and multiply it by 4. So 3 into 4 by 3, so we get 2. So we get 2 here.

And similarly, we subtract this row, this row we subtract, we subtract it from the third row.
Third row is this one, 1, 2, 0, 1, 0, 0, 3. From this we subtract. So what we will get? We are
subtracting, so this will become 0, right? This will become 0. Yeah, this will become 0 here.
And then we will have here, okay, here we have 2. We have 2 here and after dividing by 3 it
became 1 by 3. So 2 - 1 by 3 we will have. So that will be 5 by 3.

So that is 5 by 3 here. And then here what we will have? This is 0. Because 0 - 0 is 0 here, so
we have 0 here. Then in the s2 what we have? s2 column here we have 1. So 1 - 0 will remain
1, so that is 1 here. Under A 2 column what we will have? A 2 column here it is 0. And here
what we did was we divided by 3 and then after dividing by 3 we are subtracting. So this is 0.
So this will remain 0. And then what we will have here? Here we have 3. After dividing by 3
it became 1, so 3 - 1 is =2.

(Refer Slide Time: 24:21)

So we get 2 here. Now what we have? Let us find Z j. Z j =CB into x j. So - 2 into 1 is - 2. -
M into 0 is 0. 0 into 0 is 0. So total will be - 2. So this is - 2 here. Then - 2 into 1 by 3, - 2 by
3 we have. And then we have - 5 M by 3. And then 0 into 5 by 3, so we have 0. So total is - 2
- 5 M by 3, so we get this.

And then similarly, - 2 into 0 is 0. - M into - 1 is + M. 0 into 0, 0. So we get M here and then
this is 0, this is 0. This is 0, so we get 0 here. And then we have - 2 into 0, - M into 1 - M, 0
into 0, so we get - M here. Then - 2 into 1 is - 2. - M into 2 is - 2 M. And 0 into 2 is 0, so we
get - 2 into, - 2 - 2 M.

Now let us see. We find C j, C j =c j - Z j. So C j is - 2, when we subtract Z j which is also -

2, we get 0 here. And then - 1 - of this quantity, - of Z j makes it - 1 + 2 by 3. So - 1 by 3 + 5
M by 3, and then you have 0 - M, so we get - M. 0 - 0 is 0. And then we get - M and we
subtract - M. So we get - M + M. So that is we get 0 here.

And we notice now that since M is large, - 1 by 3 + 5 M by 3, this is a positive quantity. So C

j is greater than 0 in the column corresponding to x2 . And therefore this is our key column.
This is our key column. And the elements of key column are 1 by 3, 5 by 3, 5 by 3. So we
divide the elements of b column by the corresponding elements of x2 column. So 1 divided by
1 by 3 gives you 3. 2 divided by 5 by 3 gives you 6 by 5. And 2 divided by 5 by 3 again gives
you 6 by 5.

Now we see that we have to find that value here in the column corresponding to theta which
is the minimum. But 6 by 5 and 6 by 5 there is a tie here. These two values are both equal and
they are less than 3. So which row we will take as the key row? Because we want to get rid of
the artificial variable A 2, so we have to take this row as key row, this one. Because if we take
this as key row, then A 2 will be, it will be going out and we will get at the intersection of A 2,
x2 will be coming in and this will be our key element.

This will be our key element. x2 will come in place of A 2 and A 2 will go out. The column
corresponding to A 2 will be, will no longer appear now. So now this new simplex table, in the
new simplex table what we will have? We will have x 1. In place of A 2 we will have x2 , then
we will have s2 here. The coefficient of x2 is - 1, so - M will be replaced by - 1.

(Refer Slide Time: 27:58)

So let us see the new…since C j is positive under x2 column, this is not an optimal solution.
We introduce x2 and drop A 2. And then revised simplex table is like this: So x 1, x2 , s2; the
coefficient of x 1 is - 2. Coefficient of x2 is - 1, coefficient of s2 is 0. So now our basic
variables are x 1, x2 and s2. This is x 1, yeah, this is x2 . This one is s1, s2. So A 1, A 2, the two
basic variables, the columns corresponding to the artificial variables are not there. Now let us
see, we have the new simplex table. In the new simplex table we have 1, 1; okay, so we see
what we have now.

(Refer Slide Time: 28:48)

Yeah. So this is our 5 by 3. We divide this equation by 5 by 3. This row, this row we divide
by 5 by 3 to bring unity here, and then subtract, after multiplying this by 1 by 3 we subtract it
from this row. And subtract, multiplying by 5 by 3 we subtract it from this row to make this
element 0, this element 0. That is, we are doing elementary row operation. So let us first
divide this key row by 5 by 3. When we divide key row by 5 by 3, what we get?

(Refer Slide Time: 29:22)

(Refer Slide Time: 29:26)

We get this one: 0, 1, okay. And 5 by 3 we are dividing. So we will get here - 3 by 5. So that
is - 3 by 5 here. And then this one is 0. And then here 2, we are dividing by 5 by 3, so it will
become 6 by 5. So this is 6 by 5. Now we subtract this row. This row is the row which will
be…with the help of which we will make the entries, these ones: 1 by 3 and 5 by 3 as zeroes.

So what we do? We multiply it by 1 by 3 and subtract it from the this row, from this row. So
what we will have? We have 1 here. And here we have 0, so we are multiplying by 1 by 3
and subtracting. So 1 - 1 by 3, this is 0, 0 into 1 by 3 is 0, so we get 1 here. It is unaffected.
Here what we will have? This is 1 by 3. We are multiplying after making this entry as 1.

After making the key element as 1, we are multiplying by 1 by 3 and subtracting, so we get 0
there. And here what we will have? This became - 3 by 5. So we are multiplying by 1 by 3
and subtracting. So - 3 by 5 into 1 by 3 =- 1 by 5. Now this quantity we have to subtract from
0. So what we get? 0 -, - 1 by 5, which is =1 by 5, so we get 1 by 5 here.

And in the s2 column let us see what we get here. We are dividing by 3 by 5, sorry 5 by 3, so
this become 3 by 5. This become 3 by 5. No, A 2 is not there. We come to this one. So are
dividing by 5 by 3. This becomes 6 by 5. And then we are multiplying by 1 by 3 and
subtracting from this row. So we get 1 -…what we have? 1 - 6 by 5 into 1 by 3. 6 by 5 into 1
by 3, so we get 1 - 2 by 5 which gives us 3 by 5. So this is 3 by 5.

Now what we have to do? So now we have to make this entry 0. So we have to multiply this
row…we have to make this entry 0, so we have to multiply this row by 5 by 3 and subtract it
from the third row. So let us see what we will get? So here this is one. We are multiplying by
5 by 3 and subtracting. So what will happen?

This will remain 0, this will become 0 and here what we will have? 0 -, we divided it by 5 by
3, so it was - 3 by 5. - 3 by 5 into, we are multiplying by 5 by 3 and subtracting, so this, this
and this this, what we have here? 0 + 1. So this is = 1, I am getting. In the column s1 this is 1.
And in the column s2 what we will get here? This is 1, okay, we divided it by 5 by 3, we
divided it by 5 by 3, this become 0. This is 0, and so what we get here?

Now why we are multiplying by 5 by 3? We are multiplying by 5 by 3 and subtracting. So 5

by 3 when you multiply and subtract what we will get? This is 0 here. So multiplying by 5 by
3 and subtracting from 0 will give 1. So this will remain 1. And what we will have there?
This element will become how much? We will have 2 -, this was 6 by 5. 6 by 5 into 5 by 3, so
we get this 2 - 2 =0.

(Refer Slide Time: 33:55)

So this becomes 0. Now let us find Z j. Z j =- 2 into 1 - 2. - 1 into 0 is 0. 0 into 0 is 0. So total

is - 2. Then - 2 into 0 is 0. - 1 into 1 is - 1. 0 into 0 is 0. So we get - 1 here. - 2 into 1 by 5, so
we get - 2 by 5. - 1 into - 3 by 5 is 3 by 5, and we get 0 into 1, so we get 3 by 5 - 2 by 5 is 1
by 5. So we get 1 by 5 here. And then - 2 into 0 is 0. - 1 into 0 is 0. 0 into 1 is 0. So we get 0
here.
And then - 2 into, okay, so now let us see, so we have got these values. Now in the b column
what we will get? - 2 into - 3 by 5, so - 6 by 5. And then - 1 into 6 by 5, so - 6 by 5. And then
we get 0 into 0, so we get - 12 by 5. So this is - 12 by 5.

Now let us see Z j, we subtract it from c j. So - 2 - of - 2, that gives 0. - 1 - of - 1 gives 0. 0 - 1

by 5, - 1 by 5 we get. And then we have 0, - 0 is 0. Now we can see C j is ≤0 for all j. C j is
≤0 for all j.

(Refer Slide Time: 35:34)

(Refer Slide Time: 35:43)

So none of C j is positive and therefore we get the optimal solution. An optimal basic feasible
solution to the problem is therefore x 1, what is x 1? x 1 =3 by 5, so x 1 =3 by 5, x2 =6 by 5 and z
=- 12 by 5. z =- 12 by 5. So now maximum of Z dash is - 12 by 5, that is maximum value of
Z dash. So - 12 by 5. And minimum of Z is what ? Minimum of Z = - of maximum of Z dash,
so - of - 12 by 5.

This value is nothing but maximum value of Z dash because we have converted the problem
to maximization. So after C j is ≤ 0 for all j, the value of Z j gives the maximum value of Z.
so we get maximum value of Z dash is - 12 by 5, and that is equal to, I mean the minimum
value of Z is - of maximum of Z dash which is 12 by 5. Thus, we get the solution of the given
problem, where we find that x 1 is 3 by 5, x2 is 6 by 5 and the minimum value of the objective
function is 12 by 5.

(Refer Slide Time: 37:12)

Let us take another problem. Maximum value of Z =3 x 1 + 2 x2 , subject to the constraints: 2

x 1 + x2 ≤2, 3 x 1 + 4 x2 ≥12; x 1, x2 ≥0.
(Refer Slide Time: 37:25)

Now we have to express the problem in the extended form.

(Refer Slide Time: 37:31)

Since it is already a maximization problem, so it will remain as such: x 1, x2 are ≥ 0; 2 x 1 + x2

≤ 2, so this constraint is ≤ type. So we will write one, we will convert it into equality by
writing 2 x 1 + x2 + 0 into, we have to consider a slack variable, so 0 into s1 and then = 2 it
becomes.

But here we have 3 x 1 + 4 x2 +…≥ type it is, ≥ type, so we get this is s1, so we have to add a
slack variable here. 2 x 1 + x2 + s1 =2. 3 x 1 + 4 x2 ; now it is ≥ 2, so we write - s2. - s2 here, - s2.
We have to subtract the surplus variable. So - s2 =12. Now we have to add…In order to get
initial basic feasible solution, we have to consider the artificial variables also.
So we will write z =3 x 1 + 2 x2 + 0 s1 + 0 s2. And then we have to add one artificial variable
here because this is ≥type. So we get one artificial variable here, + 0 into A. So artificial
variable is given as - M penalty, so - M into A we will have. So objective function Z =3 x 1 +
2 x2 will be written as Z =3 x 1 + 2 x2 + 0 s1 + 0 s2 - M into A.

And here we will have 2 x 1 + x2 + s1 and then 0 s2. Here we will have 0 s1 and 0 A here. So
these equations will now become 2 x 1 + x2 + s1 + 0 s2 + 0 A =2 and 3 x 1 + 4 x2 + 0 s1 - s2 + 0
into A equal to, no, not 0 into A, we have to add A, artificial variable we have to add, so A,
=12. Now the objective function will be written as 3 x 1 + 2 x2 ; we will have two variables s1,
s2. One is…s 1 is slack variable, s2 is surplus variable and one artificial variable A we have,
which is given penalty - M.

Now this artificial variable comes because of the fact that the second inequality is greater or
=type. So the first inequality is ≤type. So we have 2 x 1 + x2 + s1 =2. But s2 and A are also
there, so we get 0 s2 + 0 into A, we add them. And here 3 x 1 + 4 x2 and then 0 s1 and we
subtract surplus variable s2, so - s2, + artificial variable A =12. So now here we can say x 1, x2 ,
s1, s2 and A all are ≥0.
(Refer Slide Time: 41:19)

So we get this equation, this maximum value of Z, 3 x 1 + 2 x2 +0 s1 + 0 s2 - MA, subject to

2 x 1 + x2 +s1+ 0s2+ 0 A = 2; 3x 1 + 4x2 + 0s1 - s2 + A =12; x 1, x2 ,s1, s2, and A ≥ 0.

(Refer Slide Time: 41:39)

Now we will find the initial basic feasible solution. Now s2 is not a basic variable because its
value is - 12.
(Refer Slide Time: 41:46)

We can see here, there are two equations, 1, 2 and 1, 2, 3, 4, 5 variables are there: x 1, x2 , s1, s2
and A. So three variables can be chosen arbitrarily. So if we choose x 1, x2 ; x 1 =0, x2 =0 and A
=0, if we take x 1 =0, x2 =0, A =0, then from this equation what do we get? s2 =- 12. So if
choose variables x 1, x2 and A =0, we get s2 =- 12 but s2 cannot be - 12.

(Refer Slide Time: 42:25)

s2 is, it is not a basic variable because its value is - 12. A negative quantity is not feasible, so
s2 must be prevented from appearing in the final solution. And therefore we take s2 =0, from
the initial solution, so we take s2 =0. So what we do? We consider x 1 =0, x2 =0, and s2 =0.
And when we do that, we have this, initial simplex table. C js are coefficients of x 1, x2 , s1, s2
and A in the objective function. This is CB column where now basis variables are s1 and A.
And non-basic variables are x 1, x2 , and s2; x 1, x2 , and s2. So the coefficient of s1 in the
objective function is 0, coefficient of A in the objective function is 0. We then calculate Z j. Z
j =Sigma CB a i j. 0 into 2 is 0. - M into 3 is - 3 M. The total is - 3 M. Then c 3 - of 3 M is 3
+ 3 M here. And this becomes 2 + 4 M here. Here it is 0. Here it is - M, here it is 0.

And this is 0 into 2 is 0. - M into 12 is - 12 M, so this is - 12 M. Now we can see C j is

positive in the column corresponding to x 1, in the column corresponding to x2 . But M is large,
so we can say 2 + 4 M is greater than 3 + 4 M. So this is our key column. Now here what do
we do? We divide the elements of b column by the corresponding elements of x2 column. So
1 when divides 2, what we get is 2 and 4 when divides 12 we get 3. So minimum value is 2.
So this is our key row. This is our key row.

And when this, so this is our key element. Now x2 will, s1 will be replaced by x2 . x 1 will be
outgoing variable, s2 will be incoming variable. So in place of s1 we will have x2 . And here
the coefficient of x2 will come, that is 2.

(Refer Slide Time: 44:45)

And we get this table: 2, x2 , - M, A, and then we get this table. So 2, 1, okay...
(Refer Slide Time: 44:55)

Now what we do? We divide the element…the key row elements by this key element. So
when we do that, it is 1, 1 divides these elements, so they are unaffected.

(Refer Slide Time: 45:10)

(Refer Slide Time: 45:23)

So we get the same elements. We get the same elements. And then with the help of this key
element, with the help of this element 1, we make this element 0. To make this element 0,
what we do? We have multiply this row, the elements of this row by 4 and subtract from this
row. So we have to multiply it by 4 and subtract from the elements of the second row.

Now coefficient of x 1 is 2. And here in the next coefficient of x 1 is 3. So 3 - 4, we are

multiplying by 4, so 2 into 4 is 8, so 3 - 4 is - 5. So we get - 5 here. This is 0 and this we get
as what? We get here, we multiply by 4 and subtract, so - 4 we will get. So this is - 4.

And here what we will get? We multiply it by 4 and subtract from here. So 2 into 4 is 8, 8
when subtracted from 12 gives you 4, so we get 4. And then we again calculate Z j. Z j is 4 +
5 M. This is 2 here, this is 2 + 4 M. This is M, this is - M and here we get 2 into 2 is 4, - M
into 4 is - 4 M, so we get 4 - 4 M. Now this is c j - Z j. We find c j - Z j. This is - 1 + 5 M.
This is 0, this is - 2 + 4 M. This is - M. So C j < 0 for all j.

So C j is negative and an artificial variable…but what do we notice? An artificial variable A

appears at non-zero level. You see in the b column, this is 4. So this artificial variable appears
at non-zero level, therefore there exists a pseudo optimal solution to the problem. So that is
all in this lecture. Thank you very much for your attention.