1.

Optimization is the act of obtaining the best result under given

circumstances. In design, construction and maintenance of any
engineering system, engineers have to take much technological and
managerial decision at several stages.

2. The ultimate goal of all such decision is either to minimize effort

required or to maximize the desired benefit. Since the effort
required or the benefit desired in any practical situation can be
expressed as a function of certain decision variables.

3. Optimization can be defined as a process of finding the conditions

that give the maximum or minimum value of a function.
Optimization is a part of operations research. Operations research
is a branch of
Mathematics concerned with the application of scientific methods
and techniques.

There is no single method available for solving all optimization

problems efficiently. There are two fundamental methods for
solving all optimization problem.

They are

1. Classical optimization technique.

2. Computational optimization technique.
Solving an unconstrained optimization problem by classical optimization
technique.

Problem
F(x) = (x – 1)(3 – x) is given. Find the value of x, for which F(x) has its maximum
value.

Solution:
F(x) = (x – 1)(3 – x) = 4x − x² − 3

For F(x) to be minimum or maximum at a particular value of x, dF(x)/dx or

F´(x) will become 0. dF(x)/dx = 4 − 2x = 0 or, x = 2. Whether F(x)
minimum or maximum value at x = 2 we are finding the 2nd derivative of F(x).
F̋̋ ̋(x) = d(4 − 2x)/dx = −2.
So we can say that F(x) has its maximum value at x = 2.
Solving an unconstrained optimization problem by classical optimization
technique.

Problem :
F(x) = (x – 1) (3 – x) is given. Find the value of x, for which F(x) has its maximum value.

Solution:
F(x) = (x – 1) (3 – x) = 4x − x² − 3
For F(x) to be minimum or maximum at a particular value of x, dF(x)/dx or F´(x) will
become 0. dF(x)/dx = 4 − 2x = 0 or, x = 2. Whether F(x) minimum or maximum
value at x = 2 we are finding the 2nd derivative of F(x).
F̋̋ ̋(x) = d(4 − 2x)/dx = −2.
So we can say that F(x) has its maximum value at x = -2.
Now we solve a constrained optimization problem applying the method of
direct substitution.

Problem: Find the dimensions of a box of largest volume that can be inscribed
in a sphere of unit radius.

SOLUTION
Let the origin of the Cartesian coordinate system x₁, x₂, x₃ be at the center of the sphere
and the sides of the box be 2x₁, 2x₂, and 2x₃. The volume of the box is given by
f (x₁, x₂, x₃) = 8x₁x₂x₃ (E₁)

Since the corners of the box lie on the surface of the sphere of unit radius, x₁, x₂, and
x₃ have to satisfy the constraint
x₁²+x₂²+x₃² = 1. (E₂)

This problem has three design variables and one equality constraint. Hence the
equality constraint can be used to eliminate any one of the design variables from the
objective function. If we choose to eliminate x₃, Eq. (E2) gives
x₃ = (1 − x₁ − x₂)^1/2 (E3)

Thus the objective function becomes

f (x₁, x₂) = 8x₁x₂(1 − x₁² − x₂²)^1/2 (E4)
which can be maximized as an unconstrained function in two variables.
The necessary conditions for the maximum of f give
 
∂f ∕ ∂x₁ = 8x₂[(1 − x₁² − x₂²) ^1/2 − x₁² / (1 − x₁² − x₂²) ^1/2] = 0 (E5)
∂f ∕ ∂x₂ = 8x₁[(1 − x₁² − x₂²) ^1/2 − x₂² / (1 − x₁² − x₂²) ^1/2] = 0 (E6)
 
Equations (E5) and (E6) can be simplified to obtain
1 − 2x₁² − x₂² = 0
1 − x₁² − 2x₂² = 0

From which it follows that from which it follows that x₁* = x₂* = 1/√3 and hence x₃* = 1/√3.
This solution gives
The maximum volume of the box as
fmax = 8 / 3√3

To find whether the solution found corresponds to a maximum or a minimum,

We apply the sufficiency conditions to f (x1, x2) of Eq. (E4). The second-order partial
Derivatives of f at (x₁*, x₂*) are given by-

∂²f ∕ ∂x₁² = −32√3 at (x₁*, x₂*)

∂²f ∕ ∂x₂² = −32√3 at (x₁*, x₂*) 
∂²f ∕ ∂x₁∂x₂ = −16√3 at (x₁*, x₂*)

Hence the point (x₁*, x₂*) corresponds to the maximum of f.

Lagrange multiplier method:
Problem with two variables and one equality constraint

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

If there are multiple constraints related with its variables (x₁,x₂) such a way that
substitution is not possible. Then we apply Lagrange multiplier method to optimize
f(x₁,x₂).

In this method we will construct a new function which includes the objective function,
constraint and an extra variable like this way ---
L(x₁,x₂,λ) = f(x₁,x₂) + λ g(x₁,x₂)

Where L = Lagrange function and

λ = Lagrange multiplier.

Then we will find the derivative of L(x₁,x₂,λ) with respect to x₁, x₂ ,λ respectively.
Example: Find the dimensions of a cylindrical tin (with top and bottom) made
Up of sheet metal to maximize its volume such that the total surface area is equal to
Ao = 24π.

SOLUTION
If x₁ and x₂ denote the radius of the base and length of the tin, respectively,
the problem can be stated as
maximize f (x₁, x₂) = πx₁²x₂ Subject to
2πx₁² + 2πx₁x₂ = Ao = 24π

Let, g(x₁, x₂) = 2πx₁² + 2πx₁x₂ −24π = Constraint function

The Lagrange function is

L(x₁, x₂, λ) = f (x₁, x₂) + λ g(x₁, x₂) Or
L(x₁, x₂, λ) = f (x₁, x₂) + λ (2πx₁² + 2πx₁x₂ − A0)

And the necessary conditions for the maximum of f give

∂L/∂x₁ = 2πx₁x₂ + 4πλx₁ + 2πλx₂ = 0 (E1)
∂L/∂x₂ = πx₁² + 2πλx₁ = 0 (E2)
∂L/∂λ = 2πx₁² + 2πx₁x₂− Ao = 0 (E3)
Equations (E1) and (E2) lead to
λ = −(x₁x₂ / 2x₁ + x₂) = − (1/2) x₁ that is,
x₁ = ( 1/2) x₂ (E4)
And Eqs. (E3) and (E4) give the desired solution as
 x₁* = (A0/6π)^1/2 , x₂* = (2A0/3π)^1/2, and λ* = −(A0/24π)^1/2

This gives the maximum value of f as

f * =(A³o/54 π)^ ½

If A0 = 24 π, the optimum solution becomes

x₁* = 2, x₂* = 4, λ* = −1, and f * = 16 π

To see that this solution really corresponds to the maximum of f , we apply the
sufficiency
Condition of Eq. (2.44). In this case

L₁₁ = (∂²L/∂x₁²)| (X*, λ*) = 2π x₂* + 4π λ* = 4π

L₁₂ = (∂²L/∂x₁∂x₂)|(X*, λ*) = L₂₁ = 2 π x₁* + 2 π λ*= 2π
L22 = (∂²L/∂x₂²)| (X*, λ*) = 0

g₁₁ = ∂g₁/∂x₁₂²)| (X*, λ*) = 4πx₁*+ 2π x₂* = 16 π

g₁₂ =∂g₁/∂x₂| (X*, λ*) = 2πx₁*= 4π
Thus Eq. (2.44) becomes _______
4_ − z 2_ 16_
2_ 0 − z 4_
16_ 4_ 0
_______
=0

that is,
272_2z + 192_3 = 0

This gives
z = − (12/17 ) π

Since the value of z is negative, the point (x₁*, x₂*) corresponds to the maximum of f.
Introduction

Optimization with multiple decision variables and equality constraint : Lagrange

Multipliers.
Optimization with multiple decision variables and inequality constraint : Kuhn-Tucker
(KT) conditions
KT condition: Both necessary and sufficient if the objective function is concave and each
constraint is linear or each constraint function is concave.

Kuhn Tucker Conditions: Optimization Model Consider the following optimization

problem Minimize f(X) subject to
g(X)=0 for j=1,2,…,p where the decision variable vector X = [x1,x2,…,xn]

Kuhn-Tucker conditions for X*= [x1* x2* . . . xn*] to be a local minimum are