M. Tech.

ED II SEMESTER Course Code: CED11T13 ADVANCED OPTIMIZATION TECHNIQUES LPC 3 - 3

SYLLABUS
UNIT I LINEAR PROGRAMMING: Two-phase simplex method, Big-M method, duality, interpretation, applications. UNIT II ASSIGNMENT PROBLEM: Hungarians algorithm, Degeneracy, applications, unbalanced problems, Traveling salesman problem. UNIT III CLASSICAL OPTIMIZATION TECHNIQUES: Single variable optimization with and without constraints, multi variable optimization without constraints, multi - variable optimization with constraints - method of Lagrange multipliers, Kuhn-Tucker conditions. UNIT IV NUMERICAL METHODS FOR OPTIMIZATION: Nelder Meads Simplex search method, Gradient of a function, Steepest descent method, Newtons method, type s of penalty methods for handling constraints. UNIT V GENETIC ALGORITHM (GA): Differences and similarities between conventional and evolutionary algorithms, working principle, reproduction, crossover, mutation, termination criteria, different reproduction and crossover operators, GA for constrained optimization, draw backs of GA. UNIT VI GENETIC PROGRAMMING (GP): Principles of genetic programming, terminal sets, functional sets, differences between GA & GP, random population generation, solving differential equations using GP. UNIT VII MULTI-OBJECTIVE GA: Paretos analysis, Non-dominated front, multi - objective GA, Nondominated sorted GA, convergence criterion, applications of multi-objective problems. UNIT VIII APPLICATIONS OF OPTIMIZATION IN DESIGN AND MANUFACTURING SYSTEMS: Some typical applications like optimization of path synthesis of a four-bar mechanism, minimization of weight of a cantilever beam, optimization of springs and gears, general optimization model of a machining process, optimization of arc welding parameters, and general procedure in optimizing machining operations sequence. TEXT BOOKS: 1. Jasbir S. Arora (2007), Optimization of structural and mechanical systems, 1st Edition, World Scientific, Singapore. 2. Kalyanmoy Deb (2009), Optimization for Engineering Design: Algorithms and Examples , 1st Edition, Prentice Hall of India, New Delhi, India. 3. Singiresu S. Rao (2009), Engineering Optimization: Theory and Practice, 4th Edition, John Wiley & Sons, New Delhi, India. REFERENCE BOOKS: 1. D. E. Goldberg (2006), Genetic algorithms in Search, Optimization, and Machine learning, 28th Print, AddisonWesley Publishers, Boston, USA. 2. W. B. Langdon, Riccardo Poli (2010), Foundations of genetic programming, 1st Edition, Springer, New York. 3. R. Venkata Rao, Vimal J. Savsani (2012), Mechanical Design Optimization Using Advanced Optimization Techniques, 1st Edition, Springer, New York. Other Titles

1. Hamdy A Taha (2007), Operations Research An introduction, 8th Ed, Pearson

UNIT I

LINEAR PROGRAMMING

Two-phase simplex method, Big-M method, duality, interpretation, applications.

Introduction
The first formal application of Operations Research (OR) was initiated in England during World War II, when a team of British Scientists set out to make scientifically based decisions regarding the best utilization of war material. After the war, the ideas advanced in military operations were adapted to improve efficiency and productivity in the civilian sector (Taha 2007). "Operations Research (OR) is the representation of real-world systems/problems/decisions by mathematical models together with the use of quantitative methods (algorithms) for solving such models, with a view to optimizing." A mathematical model of OR consists: Decision variables, which are the unknowns to be determined by the solution to the model. Constraints to represent the physical limitations of the system An objective function An optimal solution to the model is the identification of a set of variable values which are feasible (satisfy all the constraints) and which lead to the optimal value of the objective function. In general terms, one can regard OR as being the application of scientific methods / thinking to decision making. Underlying OR is the philosophy that: decisions have to be made; and using a quantitative (explicit, articulated) approach will lead to better decisions than using non-quantitative (implicit, unarticulated) approaches. Indeed it can be argued that although OR is imperfect it offers the best available approach to making a particular decision in many instances (which is not to say that using OR will produce the right decision). Linear Programming (LP) is an optimization method applicable for the solution of problems in which the objective function and the constraints appear as linear functions of the decision variables. The constraint equations in a linear programming problem may be in the form of equalities or inequalities (SS Rao 2009). LP is a mathematical technique designed to aid managers in allocating scarce resources (such as labor, capital, or energy) among competing activities. It reflects, in the form of a model, the organizations attempt to achieve some objective (frequently maximizing profit/rate of return, minimizing costs) in view of limited or constrained resources (available capital/labor, service levels, available machine time, capital).

The objective of a decision maker in a linear programming problem is to maximize or minimize an objective function in consideration with resources subjected to some constraints. The constraints take the form of linear inequalities, hence the name "linear" for the problem in analysis.

The general linear programming problem can be stated in the following standard forms: SCALAR FORM Minimize f (x1, x2, . . . , xn) = c1x1 + c2x2 + + cnxn subject to the constraints a11x1 + a12x2 + + a1nxn = b1 a21x1 + a22x2 + + a2nxn = b2 ... am1x1 + am2x2 + + amnxn = bm x1 0 x2 0 ... xn 0 where cj , bj , and aij (i = 1, 2, . . . ,m; j = 1, 2, . . . , n) are known constants, and xj are the decision variables.

(1)

(2)

(3)

The characteristics of a linear programming problem, stated in standard form, are 1. The objective function is of the minimization type. 2. All the constraints are of the equality type. 3. All the decision variables are nonnegative. MATRIX FORM Minimize f (X) = cTX subject to the constraints aX = b X0 Where,

(4) (5) (6)

, { } {

, } { }

, a= [ ]

Any linear programming problem can be expressed in standard form by using the following transformations. 1. The maximization of a function f (x1, x2, . . . , xn) is equivalent to the minimization of the negative of the same function. e.g: The objective function minimize f = c1x1 + c2x2 + + cnxn is equivalent to maximize f = f = c1x1 c2x2 cnxn

Consequently, the objective function can be stated in the minimization form in any linear programming problem. 2. In most engineering optimization problems, the decision variables represent some physical dimensions, and hence the variables xj will be nonnegative. However, a variable may be unrestricted in sign in some problems. In such cases, an unrestricted variable (which can take a positive, negative, or zero value) can be written as the difference of two nonnegative variables. Thus if xj is unrestricted in sign, it can be written as xj = xj xj , where xj 0 and xj 0 It can be seen that xj will be negative, zero, or positive, depending on whether x j is greater than, equal to, or less than xj 3. If a constraint appears in the form of a less than or equal to type of inequality as ak1x1 + ak2x2 + + aknxn bk it can be converted into the equality form by adding a nonnegative slack variable S1 as follows: ak1x1 + ak2x2 + + aknxn + S1= bk A linear constraint of the form can be converted into equality by adding a new, nonnegative variable to the left-hand side of the inequality. Such a variable is numerically equal to the difference between the right- and left-hand sides of the inequality and is known as slack variable. It represents the part of unutilized resource of type . Similarly, if the constraint is in the form of a greater than or equal to type of inequality as ak1x1 + ak2x2 + + aknxn bk it can be converted into the equality form by subtracting a variable as ak1x1 + ak2x2 + + aknxn S1= bk where S1 is a nonnegative variable known as a surplus variable. A linear constraint of the form can be converted into equality by adding a new, nonnegative variable to the left-hand side of the inequality. Such a variable is numerically equal to the difference between the right- and left-hand sides of the inequality and is known as surplus

variable. It represents the part of resource required in excess of the minimum limit of type

It can be seen that there are m equations in n decision variables in a linear programming problem. We can assume that m < n; for if m > n, there would be m n redundant equations that could be eliminated. The case n = m is of no interest, for then there is either a unique solution X that satisfies Eqs. (3.2 or 3.5) and (3.3 or 3.6) (in which case there can be no optimization) or no solution, in which case the constraints are inconsistent. The case m < n corresponds to an underdetermined set of linear equations, which, if they have one solution, have an infinite number of solutions. The problem of linear programming is to find one of these solutions that satisfies Eqs. (3.2 or 3.5) and (3.3 or 3.6) and yields the minimum value of f (Objective function in consideration for optimization).

LP Solutions:
When an LP is solved, one of the following four cases will occur: 1. The LP has a unique optimal solution. 2. The LP has alternative (multiple) optimal solutions. It has more than one (actually an infinite number of) optimal solutions 3. The LP is infeasible. It has no feasible solutions (The feasible region contains no points). 4. The LP is unbounded. In the feasible region there are points with arbitrarily large (in a max problem) objective function values.

PS: Please refer - SSRao (2009) for more fundamental concepts.

Two-phase simplex method