Pages From Slide - Optimization

Engineering Optimization - DMS6036
Dr.-Ing. Getachew Shunki Tibba
Mettu University
G. S. Tibba (MeU) Engineering Optimization - DMS6036 1 / 113

Outline I
1 Introduction to Optimization
Introduction
Engineering Applications of Optimization
Engineering Design Procedures
Statement of Optimization Problems
Classification of Optimization Problems
Optimization Techniques
2 Classical Optimization Techniques
Single Variable Optimization
Multi Variable Optimization with no Constraints
Multi Variable Optimization with Equality Constraints
Solution by Direct Substitution
Multi Variable Optimization with Inequality Constraints
Solution of Optimization Problems Using Python
3 Unconstrained Optimization Techniques
Outline II
Introduction to Optimum Design
General Principles of Optimization
Problem Formulation and their Classifications
Single Variable and Multivariable Optimization
Techniques of Unconstrained Minimization
Golden Section
Random, Pattern and Gradient Search Methods
Interpolation Methods
4 Constrained Optimization Techniques

Optimization with Equality and Inequality Constraints
Direct Methods
Indirect Methods using Penalty Functions
Geometric programming
5 Modern Methods of Optimization

Genetic Algorithms
Simulated Annealing
Outline III
Fuzzy Optimization
Neural-Network-Based Methods
6 Topology Optimization
Problem Formulation and Parameterization of Design
Solution Methods
Topology Optimization as a Design Tool
Combining Topology and Shape Design
Buckling Problems
Stress Constraints
7 Evolutionary Structural Optimization (ESO) Methods
ESO Based on Stress Level
Evolutionary Methods
Two-bar Frame
Michell Type Structure
ESO for Stiffness or Displacement Optimization
Outline IV
Bi-directional Evolutionary Structural Optimization (BESO) method
BESO Based on von Mises Stress
Topology Optimization for Natural Frequency
8 Static Applications
Structural Applications
Design of Simple Truss Members
Design Applications
Design of Simple Axial, Transverse Loaded Members for Minimum Cost, Weight
Design of Shafts and Torsionally Loaded Members
Design of Springs
9 Dynamic Applications
Dynamic Applications
Optimum Design of Single Degree of Freedom Systems
Optimum Design of Two Degree of Freedom Systems
Optimum Design of Vibration Absorbers
Application in Mechanisms
Outline V
Optimum Design of Simple Linkage Mechanisms

References I
1 Rao, S. S., Engineering Optimization, Theory and Practice, 3rd

edition, John Wiley Publications, New York, 1996.
2 Structural Optimization, Raphael T. Haftka and ZaferGurdal, Kluwer
Academic Publishers
3 Practical Optimization Methods with Mathematical Applications, M.
Asghar Bhatti, Springer
4 Topology Optimization Theory, Methods and Applications, M. P.
Bendse, Q. Sigmund Evolutionary Topology Optimization of
Continuum Structures, Methods and Applications, X. Huang, Y.M.
Xie, Wiley, 2010
5 Johnson Ray, C., Optimum design of mechanical elements, Wiley,
John & Sons, 1990.

References II
6 Kalyanamoy Deb, Optimization for Engineering design algorithms and

Examples, Prentice Hall of India Pvt. 1995.
7 Goldberg, D.E., Genetic algorithms in search, optimization and
machine, Barnen, AddisonWesley, New York, 1989.

Method of Delivery and Evaluation I
Method of Delivery
Lectures supported by tutorials
Group project work
Individual assignment
Evaluation
30% two journal paper study reports (15% each) with output
presentation
Mid semester examination 30%
Final examination 40%

Method of Delivery and Evaluation II
Attendance Requirement
Minimum of 75% attendance during lecture hours
100% attendance during practical work sessions, except for some
unprecedented mishaps
All exercises and project works must be submitted by the specified
dead line date

Course Objectives
Objectives
To impart knowledge on various categories of existing engineering
problems and solutions to such problems through different optimization
techniques and approaches. The course is intended to introduce the
students the techniques of optimization in Engineering Design. The course
will introduce:
Principle of optimization, formulation of objective function design
constraints and classification of optimization problem.
Optimization techniques such as Single variable and multi-variable
optimization, multi-objective optimization.
Static and dynamic applications where the optimization techniques
will be used.
Outcome
It helps the engineers to get familiarized with the different approaches of
optimizing (maximizing or minimizing) an engineering problem or a
function which is essentially required in industries today.
Introduction to Optimization Introduction
Optimization is act of obtaining the best result under given

circumstances.
Technological and managerial decisions at several stages.
To minimize the effort required or to maximize the desired benefit.
Mathematically The process of finding the conditions that give the
maximum or minimum value of a function.
f(x)
f(x)
Minimum of f(x)
x* x
Maximum of -f(x)
-f(x)
Figure 1 : Maximum and Minimum Relations of a Function

The optimum seeking methods are also known as mathematical

programming techniques and are generally studied as a part of
operations research.
Beginning of Operations Research World war II British military
allocating very scarce and limited resources to several activities.
The three Operations Research Areas

1. Mathematical programming techniques are useful in finding the
minimum of a function of several variables under a
prescribed set of constraints.
2. Stochastic process techniques can be used to analyze problems
described by a set of random variables having known
probability distributions.
3. Statistical methods enable one to analyze the experimental data and
build empirical models to obtain the most accurate
representation of the physical situation.

Table 1 : Traditional and Modern Optimization Techniques
Traditional Modern
Calculus methods Genetic algorithms
Calculus of variations Simulated annealing
Nonlinear Programming Ant colony optimization
Geometric programming Particle swarm optimization
Quadratic programming Neural networks
Linear Programming Fuzzy optimization
Dynamic programming
Integer programming
Stochastic programming
Separable programming
Multiobjective programming
Network methods: CPM and PERT
Game theory

Introduction to Optimization Engineering Applications of Optimization
There are a number of application areas of optimization in Engineering.

Few are listed below:
Design of aircraft and aerospace structures for minimum weight
Finding the optimal trajectories of space vehicles
Design of civil engineering structures such as frames, foundations,
bridges, towers, chimneys, and dams for minimum cost
Minimum-weight design of structures for earthquake, wind, and other
types of random loading
Design of water resources systems for maximum benefit
Optimal plastic design of structures
Optimum design of linkages, cams, gears, machine tools, and other
mechanical components
Selection of machining conditions in metal-cutting processes for
minimum production cost
Design of material handling equipment, such as conveyors, trucks,
and cranes, for minimum cost
Design of pumps, turbines, and heat transfer equipment for maximum
efficiency
Introduction to Optimization Engineering Design Procedures
Conceptual
Design
Initial Design
Modeling
and
Simulation
Iterative
Redesign Evaluation
No
Acceptable?
Yes
Solution
Figure 2 : Typical Iterative Design Procedure

Conceptual Design:
Conceptual first phase of design
Design
provides a description of the proposed
Initial Design system
what it should do, behave, and look
Modeling like
and drawings and solid models are used for
Simulation
the description
Iterative
Redesign Evaluation Conceptual Design can be:
Innovative conceptual design
No Selection from available concepts
Acceptable?
Modification in the design of existing
Yes systems
Solution

Conceptual
Design
Initial Design:
Initial Design Once the conceptual design is obtained
the initial design is specified in terms
Modeling of:
and
the configuration of the system,
Simulation
the given quantities from the problem
Iterative
Redesign statement, and
Evaluation
an appropriate selection of the design
variables.
No
Acceptable?
Yes
Solution

Modeling and Simulation

Modeling
Conceptual The process of simplifying a given problem
Design so that it may be represented in terms of
Initial Design
a system of equations, for analysis, or
a physical arrangement, for
Modeling
experimentation,
and
Simulation is termed modeling.
Iterative
Redesign Evaluation Simulation
Simulation is the process of subjecting a
No model for a given system to various inputs
Acceptable?
to determine how it behaves and thus
Yes
predict the characteristics of the actual
Solution
physical system.

Conceptual
Design
Initial Design Evaluation:

the simulation results are evaluated
Modeling
and
against meeting the requirements and
Simulation not violating constraints.
Iterative different designs may be considered and
Redesign Evaluation the best that is optimum will be taken.
No
Acceptable?
Yes
Solution

Introduction to Optimization Statement of Optimization Problems
An optimization or a mathematical programming problem can be

stated as follows. Find set of variables
X = {x1 , x2 , ..., xp }T which minimizes f(X)
with constraints,
gi (X) = 0 (1)
and
hj (X) Cj (2)
where
X is design vector, gi (X), (i = 1, 2, ..., m) are equality constraints and
hj (X), (j = 1, 2, ..., n) are inequality constraints.

Some optimization problems do not involve any constraints (unconstrained

optimization) and can be stated as
X = {x1 , x2 , ..., xp }T which minimizes f(X)
Design Vectors
Design variables: are set of quantities that define Engineering systems or
components.
Consider the gear pair shown in the following figure, characterized by

its face width b, number of teeth T1 and T2 , center distance d,
pressure angle , tooth profile, and material.
If d, , tooth profile, and material of the gears are fixed in advance,
these quantities can be called preassigned parameters.
The remaining quantities can be collectively represented by design
vector X = {x1 , x2 , x3 }T = {b, T1 , T2 }T .

Design Constraints
In many practical problems, the design T1
N1
variables cannot be chosen arbitrarily.
They must satisfy certain restrictions
called design constraints. d
T2
1 Behavior or functional constraints:
represent limitations on the behavior
or performance of the system. N2
Example is strength requirement.
2 Geometric or side constraints: b
represent physical limitations on
design variables, such as availability, Figure 3 : Gear Pair
fabricability, and transportability.

x2 Side Constraint g3 = 0 Behavior Constraint g2 = 0
Feasible Region
Free Point
Behavior Bound Acceptable Behavior Constraint g1 = 0

Constraint
g4 = 0 Points
Free
Unacceptable Side Constraint g5 = 0
Points
Bound Unacceptable Points

x1
Figure 4 : Example of Constraint Surfaces

Objective Functions
The purpose of optimization is to choose the best among the many
acceptable designs.
The criterion with respect to which the design is optimized, when
expressed as a function of the design variables, is known as the
criterion or merit or objective function.
Examples of objective functions and optimization criteria:
In aircraft and aerospace structural design problems: Minimization of
weight.
In civil engineering structural designs Minimization of cost.
In mechanical engineering systems design Maximization of efficiency.
A design may have more than one objective functions.
There may be cases where the optimization with respect to a
particular criterion may lead to results that may not be satisfactory
with respect to another criterion.

Objective Functions ...

If there are more than one criterion to be satisfied simultaneously, the
problem is said to be multiobjective one.
In such cases, an overall objective function is constructed as a linear
combination of the objective functions. Example:
f (X) = 1 f1 (X) + 2 f2 (X) + ... + n fn (X) (3)
where 1 , 2 , ..., n are constants whose values indicate the relative

importance of one objective function relative to the others.

The locus of all points satisfying f (X) = c = constant forms a

hypersurface in the design space, and for each value of c there corresponds
a different member of a family of surfaces.
x2
Constraint Surfaces
Objective Function Surfaces
Op c1 < c2 < ... < c7

ti m
um
Po
i nt
c1
c2 c7
c3
c6
f = c4 x1
f = c5
Figure 5 : Contours of Objective Function

Example 1
A column, carrying load P = 2500kgf , is made
up of a material that has a yield stress
kgf
y = 500 cm 2 , modulus of elasticity
kgf kgf
E = 0.85e6 cm 2 , and density = 0.0025 cm3 .
The length of the column is 250cm. The stress
induced in this column should be less than the
buckling stress as well as the yield stress. The
mean diameter of the column is restricted to lie
between 2cm and 14cm, and columns with
thicknesses outside the range 0.2cm to 0.8cm
are not available in the market. The cost of the
column includes material and construction costs
and can be taken as 5W + 2d, where W is the
weight in kilogram force and d is the mean
diameter of the column in centimeters.

The design variables are the mean diameter d and tube thickness t.
X = {x1 , x2 }T = {d, t}T
The objective function to be minimized is given by:
f (X) = 5W + 2d = 5ldt + 2d = 9.82x1 x2 + 2.0x1
The induced stress is given by

P 2500 2500
i = = =
dt dt x1 x2
The buckling stress for a pin-connected column is given by
2 EI 1 2E 2
b = = (x + x22 )
l 2 dt 8l 2 1

Thus the behavior constraints can be stated as

2500
g1 (X) = 500 0
x1 x2
2500 2 E
g2 (X) = 2 (x12 + x22 ) 0
x1 x2 8l
The side constraints are given by
2 x1 14 and 0.2 x2 0.8
The side constraints can be rewritten as
g3 (X) = 2 x1 0 and g4 (X) = x1 14 0
g5 (X) = 0.2 x2 0 and g6 (X) = x2 0.8 0

The contours of the objective function are to be plotted before

finding the optimum point for a series of values of c.
f (X) = 9.82x1 x2 + 2.0x1 = c

The objective function that optimizes this problem passes through the
intersection of the curves of the two side constraints.
To find exact values of the d and t that optimize the problem, a
Python program has been written (MATLAB is also possible)with
scipy optimize module. The optimum results are: d = 5.44cm and
t = 0.293cm.

f (X) = 9.82x1 x2 + 2.0x1 = c f (X ) = 5
g1 (X) = 2500
500 0 f (X ) = 10
x1 x2
f (X ) = 15
2500 2 E
g2 (X) = x1 x2 8l 2
(x12 + x22 ) 0
f (X ) = 20
f (X ) = 25
f (X ) = 30
f (X ) = 35
f (X ) = 40
f (X ) = 45
f (X ) = 50
g1 (X )
g2 (X )

Introduction to Optimization Classification of Optimization Problems
Classification Based on:
Constraints
Constrained optimization problem
Unconstrained optimization problem
Nature of the design variables

Static optimization problems
Dynamic optimization problems
Physical structure of the problem

Optimal control problems
Non-optimal control problems

Nature of the equations involved

Nonlinear programming problem
Geometric programming problem
Quadratic programming problem
Linear programming problem
Permissible values of the design variables

Integer programming problems
Real valued programming problems
Deterministic nature of the variables

Stochastic programming problem
Deterministic programming problem

Separability of the functions

Separable programming problems
Non-separable programming problems
Number of the objective functions

Single objective programming problem
Multiobjective programming problem

Geometric Programming
A geometric programming problem (GMP) is one in which the
objective function and constraints are expressed as posynomials in X.
Find X which minimizes
a
f (X ) = ci xj ij , ci > 0, xj > 0, i = 1, 2, ..., N0 , j = 1, 2, ..., n (4)
subject to
q
gk = aik xj ijk , aik > 0, xj > 0, k = 1, 2, ..., m

Quadratic Programming Problem

A quadratic programming problem is a nonlinear programming
problem with a quadratic objective function and linear constraints.
f (X ) = c + qi xi + Qij xi xj (5)
subject to
aij xi = bj , xi 0, i = 1, 2, ..., n, j = 1, 2, ..., m
where c, qi , Qij and bj are constants.

Optimal Control Problem

An optimal control (OC) problem is a mathematical programming
problem involving a number of stages, where each stage evolves from
the preceding stage in a prescribed manner.
It is usually described by two types of variables: the control (design)
and the state variables. The control variables define the system and
govern the evolution of the system from one stage to the next, and
the state variables describe the behaviour or status of the system in
any stage.
The problem is to find a set of control or design variables such that
the total objective function (also known as the performance index)
over all stages is minimized subject to a set of constraints on the
control and state variables.

Integer Programming Problem

If some or all of the design variables x1 , x2 , .., xn of an optimization
problem are restricted to take on only integer (or discrete) values, the
problem is called an integer programming problem.
If all the design variables are permitted to take any real value, the
optimization problem is called a real-valued programming problem.
Stochastic Programming Problem

A stochastic programming problem is an optimization problem in
which some or all of the parameters (design variables and/or
preassigned parameters) are probabilistic (non-deterministic or
stochastic).
In other words, stochastic programming deals with the solution of the
optimization problems in which some of the variables are described by
probability distributions.

Separable Programming Problem

A function f (x) is said to be separable if it can be expressed as the
sum of n single variable functions, f1 (x1 ), f2 (x2 ), ., fn (xn ), that is,
f (x) = fi (xi ), i = 1, 2, ..., n (6)
where summation convention applies over indices.

A separable programming problem is one in which the objective
function and the constraints are separable and can be expressed in
standard form as:
Find x which minimizes, f (x) = fi (xi ), i = 1, 2, ..., n
subject to gj (x) = gij (xi ) bj , j = 1, 2, ..., m where bj is constant.

Multi-objective Programming Problem

A multiobjective programming problem can be stated as follows:
Find x which minimizes f1 (x), f2 (x), ..., fk (x)
subject to gj (x) 0, j = 1, 2, ..., m
where f1 , f2 , ..., fk denote the objective functions to be minimized
simultaneously.

Introduction to Optimization Optimization Techniques
Example 2
Consider the slider-crank mechanism shown in Figure 6 with the crank
rotating at a constant angular velocity . Use a graphical procedure to
find the lengths of the crank and the connecting rod to maximize the
velocity of the slider at a crank angle of = 30 for = 100rad/s. The
mechanism has to satisfy Groshofs criterion l 2.5r to ensure 360
rotation of the crank. Additional constraints on the mechanism are given
by 0.5 r 10, 2.5 l 25 and 10 x 20.
Crank, length=r
Connecting rod, length=l

Slider

Figure 6 : Slider-Crank Mechanism

Classical Optimization Techniques Single Variable Optimization
Single variable optimization methods make use of the techniques of

differential calculus.
Since some of the practical problems involve objective functions that
are not continuous and/or differentiable, the classical optimization
techniques have limited scope in practical applications.
f (x) has a relative or local minimum at x = x if f (x ) f (x + h)
for all sufficiently small positive and negative values of h.
x is called a relative or local maximum if f (x ) f (x + h) for all
values of h sufficiently close to zero.
f (x) is said to have a global or absolute minimum at x if
f (x ) f (x) for all x in the domain x will be a global maximum of
f (x) if f (x ) f (x) for all x in the domain.


A1 , A2 , A3 = Relative Maxima, f = 0, f < 0
A2 = Global Maximum
f (x) A2
B1 , B2 = Relative Minima, f = 0, f > 0
A3
B1 = Global Minimum
A1
B2
B1
a b x
a)

f does not exist
f (x) f (x) but f (x ) is local minimum
Stationary

Point, f (x) = 0
No optima m m+
o x
x x
b) c)
Figure 7 : Conditions for Relative and Global Optima

Example 3
Determine the maximum and minimum values of the function
f (x) = 12x 5 45x 4 + 40x 3 + 5
50
f (x)
40
Point 1:

f (x) = 0 30
Inflection Point
Point 2: 20

f (x) = 0
f (x)
Local Maximum 10 2
1
Point 3: 0

f (x) = 0
Local Minimum -10
3
-20
-30
-1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5
x
rth Differential of f at x
r f (x )
d r f (x ) = hi hj ...hk (7)
xi xj ...xk
Taylor series expansion of f at x

1 2 1
f (x) = f (x )+df (x )+ d f (x )+...+ d n f (x )+Rn (x , h)
2! n!
(8)
1 n+1 f
where Rn = (n+1)! d (x + h), 0 1, h = x x
Equation (7) is inserted in the Taylor Series expansion equation (8).

Example 4
Find the second order Taylors series approximation of the function
f (x1 , x2 , x3 ) = x22 x3 + x1 e x3
about the point x = (1, 0, 2).

Solution
The second order Taylors series approximation of the function f about
point x is given by
1 2
f (x) = f (1, 0, 2) + df (1, 0, 2) + d f (1, 0, 2)
2!
f (x) = e 2 + [(h1 + h3 )e 2 ] + [4h22 + (h32 + 2h1 h3 )e 2 ]
Where h1 = x1 1, h2 = x2 , and h3 = x3 + 2

Classical Optimization Techniques Multi Variable Optimization with no Constraints
Necessary condition
If f (x) has an extreme point (maximum or minimum) at x = x and if the
first partial derivatives of f (x) exist at x , then
f (x ) f (x ) f (x )
= = ... = =0
x1 x2 xn
Sufficient condition
A sufficient condition for a stationary point x to be an extreme point is
that the matrix of second partial derivatives (Hessian matrix) of f (x )
evaluated at x is
1 Positive definite when x is a relative minimum point
2 Negative definite when x is a relative maximum point
Tests for positive definiteness of matrix A:
1 Eigenvalue test: 1 > 0, 2 > 0, ..., n > 0.
2 Scalar xT Ax > 0 for x > 0
Example 5
The following figure shows two frictionless rigid bodies (carts) A and B
connected by three linear elastic springs having spring constants k1 , k2 ,
and k3 . The springs are at their natural positions when the applied force P
is zero. Find the displacements x1 and x2 under the force P by using the
principle of minimum potential energy.
k1 B
A
k2 k3
P
x1 x2

According to the principle of minimum potential energy, the system

will be in equilibrium under the load P if the potential energy is a
minimum.
The potential energy, U, of the system is the difference between
strain energy of springs and work done by external forces

1 2 1 2 1 2
U= k2 x1 + k3 (x2 x1 ) + k1 x2 Px2
2 2 2
The necessary condition for the minimum of U are

U U
= k2 x1 k3 (x2 x1 ) = 0, and = k3 (x2 x1 ) + k1 x2 P = 0
x1 x2
Pk3 P(k2 + k3 )
x1 = , and x2 =
k1 k2 + k1 k3 + k2 k3 k1 k2 + k1 k3 + k2 k3

The sufficiency conditions for the minimum at (x1 , x2 ) can also be

verified by testing the positive definiteness of the Hessian matrix of
U. The Hessian matrix of U evaluated at (x1 , x2 ) is:
2
U 2U
x 2 x 1 x 2 k 2 + k3 k 3
J|(x1 ,x2 ) = 2 U
1
2U
=
x x 2
k3 k1 + k3
1 2 x2
(x1 ,x2 )
The determinants of the square submatrices of J are
J1 = |k2 + k3 |= k2 + k3 > 0

k2 + k3 k3
J2 = = k1 k2 + k1 k3 + k2 k3 > 0
k3 k1 + k3
since the spring constants are always positive. Thus the matrix J is
positive definite and hence (x1 , x2 ) corresponds to the minimum of
potential energy.

Semi-definite case
The sufficient conditions for the case when the Hessian matrix of the
given function is semi-definite:
In case of a function of a single variable, the higher order derivatives in the

Taylors series expansion are investigated
The sufficient conditions for a function of several variables for the
case when the Hessian matrix of the given function is semi-definite:
Let the partial derivatives of f of all orders up to the order k 2 be

continuous in the neighborhood of a stationary point x , and
d r f |x=x = 0 where 1 r k 1
d k f |x=x 6= 0
so that d k f |x=x is the first nonvanishing higher-order differential of f at
x .
If k is even:
x is a relative minimum if d k f |x=x is positive definite
x is a relative maximum if d k f |x=x is negative definite
If d k f |x=x is semidefinite, no general conclusions can be drawn
If k is odd, x is not an extreme point of f (x ).
Saddle Point
In the case of a function of two variables f (x, y ), the Hessian matrix
may be neither positive nor negative definite at a point (x , y ) at
which
f f
= =0 (9)
x y
In such a case, the point (x*,y*) is called a saddle point.
The characteristic of a saddle point is that it corresponds to a relative
minimum or maximum of f (x, y ) wrt one variable, say, x (the other
variable being fixed at y = y ) and a relative maximum or minimum
of f (x, y ) wrt the second variable y (the other variable being fixed at
x ).
Example 6
Consider the function f (x, y ) = x 2 y 2 . For this function:
f f
= 2x, = 2y
x y
These first derivatives are zero at x = 0 and y = 0. The Hessian matrix
of f at (x , y ) is given by:

2 0
J=

0 2
Since this matrix is neither positive definite nor negative definite, the point
(x = 0, y = 0) is a saddle point.

It can be seen from the figure that f (x, y ) = f (x, 0) has a relative
minimum and f (x , y ) = f (0, y ) has a relative maximum at the saddle
point (x , y ).
f (x, y )
(x , y )
Figure 8 : Saddle Point for the function f (x, y ) = x 2 y 2

f(x,y)=x^2-y^2
2
2.0
1.5
1.0
0.5
0.0
2.0 0.5 y
1.5
1.0
0.5 1.0
0.0
0.5 1.5
1.0
x 1.5 2.0
2.0
Figure 9 : Contour of the function f (x, y ) = x 2 y 2

Example 7
Find the extreme points of the function
f (x, y ) = x 3 + y 3 + 2x 2 + 4y 2 + 6
The necessary conditions for the existence of an extreme point are:

f
= 3x 2 + 4x = x(3x + 4) = 0
x
f
= 3y 2 + 8y = y (3y + 8) = 0
y
These equations are satisfied at the points:
(0, 0), (0, 8/3), (4/3, 0), and(4/3, 8/3)

To find the nature of these extreme points, we have to use the sufficiency
conditions. The second order partial derivatives of f are given by:
2f 2f
= 6x + 4, =0
x 2 xy
2f 2f
= 0, = 6y + 8
y x y 2
The Hessian matrix of f is given by:

6x + 4 0
J=
0 6y + 8
Point x J1 J2 Nature of J Nature of x f (x)

(0, 0) 4 32 Pos. definite Rel. minimum 6
(0, 8/3) 4 -32 Indefinite Saddle point 418/27
(4/3, 0) -4 -32 Indefinite Saddle point 194/27
(4/3, 8/3) -4 32 Neg. definite Rel. maximum 50/3
20
18
16
f(x,y)
14
12
10
1.0
0.5
0.0
0.5
2.0
1.0
1.5
1.0 1.5 y
0.5 2.0
0.0 2.5
x
0.5
1.0 3.0
Figure 10 : Saddle Point for the function f (x, y ) = x 3 + y 3 + 2x 2 + 4y 2 + 6

Classical Optimization Techniques Multi Variable Optimization with Equality Constraints
Minimize f (x) subject to gj (x) = 0, j = 1, 2, ..., m where
x = {x1 , x2 , ..., xn } (10)
Here m n, otherwise the problem becomes over defined and, in

general, there will be no solution.
The problem can be solved using one of the following methods:
direct substitution
the method of constrained variation
the method of Lagrange multipliers

Solution by Direct Substitution

For a problem with n variables and m equality constraints:
Solve the m equality constraints and express any set of m variables in
terms of the remaining n m variables
Substitute these expressions into the original objective function, the
result is a new objective function involving only n m variables
The new objective function is not subjected to any constraint, and
hence its optimum can be found by using the unconstrained
optimization techniques.
Simple in theory
Not convenient from a practical point of view as the constraint
equations will be nonlinear for most of the problems
Suitable only for simple problems

Example 8
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 ) = 8x1 x2 x3
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

This problem has three design variables and one equality constraint. Let us
arbitrarily eliminate x3 :
x3 = (1 x12 x22 )1/2
Thus, the objective function becomes:
f (x1 , x2 ) = 8x1 x2 (1 x12 x22 )1/2
The necessary conditions for the maximum of f give:
x12

f 2 2 1/2
= 8x2 (1 x1 x2 ) =0
x1 (1 x12 x22 )1/2
x22

f 2 2 1/2
= 8x1 (1 x1 x2 ) =0
x2 (1 x12 x22 )1/2
which can be simplified as: 1 2x12 x22= 0 and 1 x12 2x22= 0
From which it follows that x1 = x2 = 1/ 3 and hence x3 = 1/ 3
This solution gives the maximum volume of the box as:

8
fmax =
3 3
To find whether the solution found corresponds to a maximum or
minimum, we apply the sufficiency conditions to f (x1 , x2 ) of the
equation f (x1 , x2 ) = 8x1 x2 (1 x12 x22 )1/2 . The second order partial
derivatives of f at (x1 , x2 ) are given by:
2f 8x1 x2
=
x12 (1 x12 x22 )1/2
x13

8x2 2 2 1/2
+ 2x1 (1 x1 x2 )
1 x12 x22 (1 x12 x22 )1/2
32
=
3

The second order partial derivatives of f at (x1 , x2 ) are given by:

2f 8x1 x2
2
=
x2 (1 x12 x22 )1/2
x23

8x1 2 2 1/2
+ 2x2 (1 x1 x2 )
1 x12 x22 (1 x12 x22 )1/2
32
=
3
Similarly,
2f 8x22
= 8(1 x12 x22 )1/2
x1 x2 (1 x12 x22 )1/2
8x12 x22

2 2 1/2
(1 x1 x2 ) +
1 x12 x22 (1 x12 x22 )1/2
16
=
3
The Hessian matrix of f is negative definite at (x1 , x2 ). Hence the point
(x1 , x2 ) corresponds to the maximum of f .
Solution by Constrained Variation

Let us start with problem with two design variables,
Minimize f (x1 , x2 ), subject to g (x1 , x2 ) = 0
A necessary condition for f to have a minimum at some point (x1 , x2 )
is that the total derivative of f (x1 , x2 ) with respect to x1 must be
zero at (x1 , x2 )
f f
df = dx1 + dx2 = 0 (11)
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
g (x1 + dx1 , x2 + dx2 ) = 0 (12)
Taylors series expansion of equation (12) gives us,
g g
dg = dx1 + dx2 = 0 (13)
x1 x2
A necessary condition for f in order to have (x1 , x2 ) as an extreme

point (minimum or maximum)

f g f g
=0 (14)
x1 x2 x2 x1 (x ,x )
1 2
Example 9
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.

a
2y
x, Neutral Axis
x 2 + y 2 = a2
2x
Figure 11 : Log Cross Section

The tensile stress induced in a rectangular beam at any fiber

located at a distance y from the neutral axis is given by
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 the depth of the rectangular beam shown in the
figure are 2x and 2y , respectively, the maximum tensile stress induced
is given by:
3 M
=
4 xy 2
3 M
The problem has objective function f (x1 , x2 ) = 4 x1 x22 subjected to
constraint, x12 + x22 = a2 or g (x1 , x2 ) = x12 + x2 a 2 = 0
2

The following relation can be used to get the extreme points:

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

which gives us, x2 = 2x1
Back substitution in the constraint equation gives, x1 = a and
3
x2 = 2 a3

Pages From Slide - Optimization

Diunggah oleh

Informasi Dokumen

Judul Asli

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

Pages From Slide - Optimization

Diunggah oleh

Hak Cipta:

Format Tersedia

Engineering Optimization - DMS6036

Dr.-Ing. Getachew Shunki Tibba

G. S. Tibba (MeU) Engineering Optimization - DMS6036 1 / 113

4 Constrained Optimization Techniques

5 Modern Methods of Optimization

G. S. Tibba (MeU) Engineering Optimization - DMS6036 6 / 113

1 Rao, S. S., Engineering Optimization, Theory and Practice, 3rd

G. S. Tibba (MeU) Engineering Optimization - DMS6036 7 / 113

6 Kalyanamoy Deb, Optimization for Engineering design algorithms and

G. S. Tibba (MeU) Engineering Optimization - DMS6036 8 / 113

G. S. Tibba (MeU) Engineering Optimization - DMS6036 9 / 113

G. S. Tibba (MeU) Engineering Optimization - DMS6036 10 / 113

Optimization is act of obtaining the best result under given

Figure 1 : Maximum and Minimum Relations of a Function

The optimum seeking methods are also known as mathematical

The three Operations Research Areas

G. S. Tibba (MeU) Engineering Optimization - DMS6036 13 / 113

Table 1 : Traditional and Modern Optimization Techniques

G. S. Tibba (MeU) Engineering Optimization - DMS6036 14 / 113

There are a number of application areas of optimization in Engineering.

Figure 2 : Typical Iterative Design Procedure

G. S. Tibba (MeU) Engineering Optimization - DMS6036 16 / 113

G. S. Tibba (MeU) Engineering Optimization - DMS6036 17 / 113

G. S. Tibba (MeU) Engineering Optimization - DMS6036 17 / 113

Modeling and Simulation

G. S. Tibba (MeU) Engineering Optimization - DMS6036 17 / 113

Initial Design Evaluation:

G. S. Tibba (MeU) Engineering Optimization - DMS6036 17 / 113

An optimization or a mathematical programming problem can be

X = {x1 , x2 , ..., xp }T which minimizes f(X)

G. S. Tibba (MeU) Engineering Optimization - DMS6036 18 / 113

Some optimization problems do not involve any constraints (unconstrained

X = {x1 , x2 , ..., xp }T which minimizes f(X)

Consider the gear pair shown in the following figure, characterized by

G. S. Tibba (MeU) Engineering Optimization - DMS6036 19 / 113

G. S. Tibba (MeU) Engineering Optimization - DMS6036 20 / 113

x2 Side Constraint g3 = 0 Behavior Constraint g2 = 0

Behavior Bound Acceptable Behavior Constraint g1 = 0

Bound Unacceptable Points

Figure 4 : Example of Constraint Surfaces

G. S. Tibba (MeU) Engineering Optimization - DMS6036 21 / 113

G. S. Tibba (MeU) Engineering Optimization - DMS6036 22 / 113

Objective Functions ...

f (X) = 1 f1 (X) + 2 f2 (X) + ... + n fn (X) (3)

where 1 , 2 , ..., n are constants whose values indicate the relative

G. S. Tibba (MeU) Engineering Optimization - DMS6036 23 / 113

The locus of all points satisfying f (X) = c = constant forms a

Op c1 < c2 < ... < c7

Figure 5 : Contours of Objective Function

G. S. Tibba (MeU) Engineering Optimization - DMS6036 24 / 113

G. S. Tibba (MeU) Engineering Optimization - DMS6036 25 / 113

X = {x1 , x2 }T = {d, t}T

The objective function to be minimized is given by:

f (X) = 5W + 2d = 5ldt + 2d = 9.82x1 x2 + 2.0x1

The induced stress is given by

G. S. Tibba (MeU) Engineering Optimization - DMS6036 26 / 113

Thus the behavior constraints can be stated as

2 x1 14 and 0.2 x2 0.8

The side constraints can be rewritten as

g3 (X) = 2 x1 0 and g4 (X) = x1 14 0

g5 (X) = 0.2 x2 0 and g6 (X) = x2 0.8 0

G. S. Tibba (MeU) Engineering Optimization - DMS6036 27 / 113

The contours of the objective function are to be plotted before