Optimization Techniques- 20CYS201

Dr. Satyendra Singh Chauhan

Assistant Professor, Department of Mathematics,
ASE, Amrita Vishwa Vidyapeetham, Chennai- 601103
[email protected], c [email protected]
 Lecture- 5

Dr. Satyendra (ASE, Chennai) Optimization Techniques- 20CYS201 2 / 17

Exercise

Ex. 1
Minimize the function

f (x1 , x2 , x3 ) = 9 − 8x1 − 6x2 − 4x3 + 2x12 + 2x22 + x32 + 2x1 x2 + 2x1 x3

subject to the constraint x1 + x2 + 2x3 = 3 by direct substitution method.

Ex. 2
Find the values of x, y, and z that minimize the function
6xyz
f (x, y, z) =
x + 2y + 2z
when x, y, and z are restricted by the relation xyz = 16.

Dr. Satyendra (ASE, Chennai) Optimization Techniques- 20CYS201 3 / 17

Solution by Method of Constrained Variation

Objective
Minimize the function f (x1 , x2 ) subject to the constraint g(x1 , x2 ) = 0.

A necessary condition for the function f (x1 , x2 ) to have a minimum at some

point (x1∗ , x2∗ ) is that the total derivative of function f (x1 , x2 ) with respect to x1
must be zero. The total derivative of the function f (x1 , x2 ) is given by

∂f ∂f
df = dx1 + dx2 = 0
∂ x1 ∂ x2

Since g(x1∗ , x2∗ ) = 0 at the minimum point, any variations dx1 and dx2 taken
about the point (x1∗ , x2∗ ) are called admissible variations provided that the new
point lies on the constraint g(x1∗ + dx1 , x2∗ + dx2 ) = 0.
The Taylor’s series expansion of the function in the above equation about the
point (x1∗ , x2∗ ) gives

Dr. Satyendra (ASE, Chennai) Optimization Techniques- 20CYS201 4 / 17

 ∂g ∗ ∗ ∂g ∗ ∗
g(x1∗ + dx1 , x2∗ + dx2 ) ≃ g(x1∗ , x2∗ ) + (x1 , x2 )dx1 + (x , x )dx2 = 0,
∂ x1 ∂ x2 1 2
where dx1 and dx2 are assumed to be small. Since g(x1∗ , x2∗ ) = 0, then

∂g ∂g
dg = dx1 + dx2 at (x1∗ , x2∗ )
∂ x1 ∂ x2
∂g
Assuming that ∂ x2 ̸= 0, then
∂g
∂ x1
dx2 = − ∂g
(x1∗ , x2∗ )dx1
∂ x2

Putting the value of dx2 in the equation df , we get

 
∂f ∂ g/∂ x1 ∂ f
df = − dx1 = 0
∂ x1 ∂ g/∂ x2 ∂ x2 (x∗ ,x∗ )
1 2

Dr. Satyendra (ASE, Chennai) Optimization Techniques- 20CYS201 5 / 17

The expression on the left hand side is called the constrained variation of the
function f . Note that the equation has to be satisfied for all values of dx1 .
Since dx1 can be chosen arbitrary, then the above equation leads to
 
∂f ∂g ∂f ∂g
− =0
∂ x1 ∂ x2 ∂ x2 ∂ x1 (x∗ ,x∗ )
1 2

or
 
 ∂f ∂f 
 ∂ x1 ∂ x2 
 ∂g ∂ g  =0
∂x ∂ x2 (x∗ ,x∗ )
1
1 2

The above equation represents a necessary condition in order to have (x1∗ , x2∗ )
as an extreme point (maximum or minimum).

Dr. Satyendra (ASE, Chennai) Optimization Techniques- 20CYS201 6 / 17

Example

A beam of uniform rectangular cross section is to be cut from a log having a

circular cross section of diameter 2a. The beam has to be used as a cantilever
beam (the length is fixed) to carry a concentrated load at the free end. Find the
dimensions of the beam that correspond to the maximum tensile (bending)
stress carrying capacity.

Solution: From elementary strength of materials, we know that the tensile

stress induced in a rectangular beam (σ ) at any fiber located a distance y from
the neutral axis is given by
σ M
=
y I
where M is the bending moment acting and I is the moment of inertia of the
cross- section about the x− axis. If the width and depth of the rectangular
beam are 2x and 2y, respectively, the maximum tensile stress induced is given
by
Dr. Satyendra (ASE, Chennai) Optimization Techniques- 20CYS201 7 / 17
 M My 3M
σmax = y= 1
=
I 12 (2x)(2y)
3 4xy2
Thus for any specified bending moment, the beam is said to have maximum
tensile stress carrying capacity if the maximum induced stress (σmax ) is a
minimum. Hence we need to minimize k/xy2 or maximize Kxy2 , where
k = 3M/4 and K = 1/k, subject to the constraint x2 + y2 = a2 . Now,
k
f (x, y) = 2
xy
g(x, y) = x2 + y2 − a2
Next, we need to calculate the partial derivatives of functions f (x, y) and
g(x, y) with respect to the variables x and y
∂f k ∂g
=− 2 2 = 2x
∂x x y ∂x
∂f 2k ∂g
=− 3 = 2y
∂y xy ∂y
Dr. Satyendra (ASE, Chennai) Optimization Techniques- 20CYS201 8 / 17
Hence from the above condition
 
∂f ∂g ∂f ∂g
− =0
∂ x ∂ y ∂ y ∂ x (x∗ ,y∗ )
k 2k
− (2y) + (2x) = 0 at (x∗ , y∗ )
x 2 y2 xy3
√
=⇒ y∗ = 2x∗
√
Thus the beam of maximum tensile stress carrying capacity has a depth of 2
times its breadth. The optimum values of x and y can be obtained
a
x∗ = √
3
√
2a
y∗ = √
3

Dr. Satyendra (ASE, Chennai) Optimization Techniques- 20CYS201 9 / 17

 Lecture- 6

Dr. Satyendra (ASE, Chennai) Optimization Techniques- 20CYS201 10 / 17

Conditions for General Problem

 
 ∂f ∂f ∂f ∂f
··· ∂f 
 ∂ xk ∂ x1 ∂ x2 ∂ x3 ∂ xm 
 ∂ g1 ∂ g1 ∂ g1 ∂ g1 ∂ g1 
   ∂ xk ∂ x1 ∂ x2 ∂ x3 ··· ∂ xm 
f , g1 , g2 , g3 , ..., gm  ∂g ∂ g2 ∂ g2 ∂ g2 ∂ g2 
J =  ∂ xk2 ∂ x1 ∂ x2 ∂ x3 ··· ∂ xm  =0
xk , x1 , x2 , x3 , ..., xm  . .. .. .. .. .. 
 .. . . . . . 

 ∂ gm ∂ gm ∂ gm ∂ gm ∂ gm 
 ∂x
k ∂ x1 ∂ x2 ∂ x3 ··· ∂x m

where k = m + 1, m + 2, ..., n.
It is to be noted that the variations of the first m variables (dx1 , dx2 , ..., dxm )
have been expressed in terms of the variations of the remaining n − m
variables (dxm+1 , dxm+2 , ..., dxn ) in deriving the above equation. This implies
that the following relation is satisfied.
 
g1 , g2 , g3 , ..., gm
J ̸= 0
x1 , x2 , x3 , ..., xm
Dr. Satyendra (ASE, Chennai) Optimization Techniques- 20CYS201 11 / 17
Example
Minimize
1
f (y) = (y21 + y22 + y23 + y24 ) (0.1)
2
subject to

g1 (y) = y1 + 2y2 + 3y3 + 5y4 − 10 = 0 (0.2)

g2 (y) = y1 + 2y2 + +5y3 + 6y4 − 15 = 0 (0.3)

Solution- Since n = 4 and m = 2, we have to select two variables as

independent variables. First we show that any arbitrary set of variables cannot
be chosen as independent variables since the remaining (dependent) variable
have to satisfy the condition of eqn
 
g1 , g2 , g3 , ..., gm
J ̸= 0
x1 , x2 , x3 , ..., xm

Dr. Satyendra (ASE, Chennai) Optimization Techniques- 20CYS201 12 / 17

In term of the nation of our equation, let us take the independent variable as
x3 = y3 and x4 = y4 so that x1 = y1 and x2 = y2
Then the Jacobian of becomes
   ∂ g1 ∂ g1   
g1 , g2  ∂ y1 ∂ y2  1 2
J =  ∂ g2 ∂ g2  =  =0
y1 , y2  ∂y ∂y  1 2
1 2

And hence the necessary conditions cannot be aplied.

Next let us take independent variables as x3 = y2 and x4 = y4 so that x1 = y1
and x2 = y3 . Then the Jacobian in becomes
   ∂ g1 ∂ g1   
g1 , g2  ∂ y1 ∂ y3  1 3
J =  ∂ g2 ∂ g2  =  = 2 ̸= 0
y1 , y3  ∂y ∂y  1 5
1 3

and hence necessary condition can be aplied.

For k = m + 1 = 3

Dr. Satyendra (ASE, Chennai) Optimization Techniques- 20CYS201 13 / 17

  ∂f ∂f ∂f 
  ∂f ∂f ∂f 
  

 ∂ x3 ∂ x1 ∂ x2 

 ∂ y2 ∂ y1 ∂ y3  y2 y1 y3 
=  ∂∂ gy21 ∂ g1 ∂ g1 
 ∂ g1 ∂ g1 ∂ g1 
 
=  2 1 3  = 2y2 − 4y1 = 0 (0.4)

 ∂ x3
∂g ∂ x1 ∂ x2  ∂ y1 ∂ y3 
∂ g2 ∂ g2 
∂g ∂ g2 ∂ g2  2 1 5
 2  2
∂ x3 ∂ x1 ∂ x2 ∂ y2 ∂ y1 ∂ y3

and for k − m + 2 = n = 4
 ∂f ∂f ∂f 
  ∂f ∂f ∂f 
  
 
 ∂ x4 ∂ x1 ∂ x2   ∂ y4 ∂ y1 ∂ y3  y4 y1 y3 
 ∂ g1 ∂ g1 ∂ g1   ∂ g1 ∂ g1 ∂ g1 
 
 ∂ x4 ∂ x1 ∂ x2  =  ∂ y4 ∂ y1 ∂ y3  =  5 1 3 
∂g ∂g ∂g  ∂g ∂ g2 ∂ g2  6 1 5
 2 2 2  2
∂ x4 ∂ x1 ∂ x2 ∂ y4 ∂ y1 ∂ y3
= y4 (5 − 3) − y1 (25 − 18) + y3 (5 − 6)
= 2y4 − 7y1 − y3 = 0 (0.5)

Dr. Satyendra (ASE, Chennai) Optimization Techniques- 20CYS201 14 / 17

Equations give the necessary conditions for the minimum or maximum of f as
1
y1 = y2
2 (0.6)
7
y3 = 2y4 − 7y1 = 2y4 − y2
2
When equation (0.6) are substituted in eqns (0.2) and (0.3), we have

−8y2 + 11y4 = 10
(0.7)
−15y2 + 16y4 = 15

Solving eqn (0.7) for y2 and y4 , we have

5 155
y∗1 = − y∗3 =
74 74
5 30
y∗2 = − y∗4 =
37 37

Dr. Satyendra (ASE, Chennai) Optimization Techniques- 20CYS201 15 / 17

 The End

Dr. Satyendra (ASE, Chennai) Optimization Techniques- 20CYS201 16 / 17

 Thank You

Dr. Satyendra (ASE, Chennai) Optimization Techniques- 20CYS201 17 / 17