Sei sulla pagina 1di 14

Concept Map Introduction to Operation Research

Problem definition Model construction Model solution Model validation Result implementation

System oriented Inter-disciplinary team approach Quantitative solution

Steps Physical Mathematical Models by nature of environment Models by extent of generality

Features The use of Scientific methods to provide criteria for decisions regarding man, machine and systems involving repetitive operations

Models

Introduction to Operation Research

Definition

Scope

Phases

Tools Linear Programming Inventory Control methods Goal programming Queuing model Transportation model

Defence operations Industry Planning Agriculture Hospitals Transport Research and development

Judgement Research Action

Illustrate Linear Programming

Finance Production and operations management Human resources Marketing Distribution

Optimum utilisation of productive resources Quality of decision improved Practically applicable solutions

General Canonical

Applications Linearity Deterministic Additivity Divisibility

Advantages

Forms Development of a production schedule Establishment of an investment portfolio Selection of the product mix in a factory

Assumptions

Linear Programming

Examples

Limitations In real life situations constraints and objective function are not linear No guarantee of integer valued solution It deals with single objective

Requirements Decision variables and their relationship Well defined objective function Existence of alternative course of action Non-negative conditions on decision variables

Illustrate Graphical Analysis of Linear Programming Problems


Convexity property Boundaries of the feasible regions are lines or planes Objective function can be represented by a line Infeasible region Properties Feasible region Convex Half pane Convex polygon Redundant constraint Basic solution Basic feasible solution Optimum feasible solution Definitions

An LPP can be solved graphically when there are two decision variables

Analysis

Steps

Graphical Analysis of Linear Programming Problems

1. Formulate 2. Plot

Method Exceptional cases

3. Identify 4. Determine

Working rule Classification of the feasible point

A unique optimum solution Multiple optimal solution An unbounded solution Infeasible solution

5. Identify

Illustrate the Simplex method


Initial feasible solution of LPP Solution of LPP in Canonical form by simplex method 1. Locate the negative number 2. Form ratios 3. User elementary row operations 4. Replace variable in pivot rows and standard column 5. Repeat step1 through 4 6. Arrive at an optimal solution

All constraints are equations All variables are non-negative The objective function is of maximisation type

Includes Characteristics

Solution of LPP in standard form Algorithm steps

Solution of LPP for artificial variables Big-M method

Simplex Method

Standard form of LPP

Equation Given by

Solution of LPP when constant is large Two phase method Maximise or minimise Z= subject to i=1,2,m and 0, j=1,2,.n 0, i=1,2.m

Steps Break the last row into two rows Follow step 1 through 4 of simplex method Artificial variable ceases to be basic variable Delete the last row in the table Arrive at optimal solution

Steps Phase I Phase II

Illustrate the Duality in Linear Programming Problem


Values of linear programming parameters might change Linear programming parameters have an uncertainty factor attached to them

Importance of Sensitivity Analysis When input > output; System=Disequilibrium When input=output; System=Equilibrium

Economic Interpretation

Duality in Linear Programming Problem

Importance

Reduce the labour of computation Useful in planning future decisions

Guidelines

Maximisation in the primal becomes minimisation in the dual and vice-versa. () in the primal becomes () in the dual and vice-versa. Co-efficients in the objective function of the primal become constants in the objective function of the dual. Constants in the constraints of the primal become coefficients in the constraints of the dual. If primal has n variables and m constraints then dual will have m variables and n constraints. Variables in both primal and dual are non-negative.

Transportation Problem Unit-6


Transportation Problem (TP)
Formulation If the number of basic cells < maximum number of positive basic variables Degeneracy in transportation problem Developed during Case 1 Initial assignment Case 2 Iteration stage Formulation of TP Algorithm used Transportation Algorithm (MODI method) Step 4 Repeating steps 2 and 3 until the optimal solution is obtained Step 3 Improving the solution when it is optimal Finding an initial basic feasible solution Step 1 Step 2 Testing the solution for optimality Techniques Step 1 Step 2 Maturity period Step 3 Methods North-west corner rule

Matrix minimum method or least cost method Vogels approximation Maturity method period Determine the net evaluations for the Maturity period non-basic variables (empty cells) Determine the Maturity period entering variable Determine the leaving variable Compute a better basic feasible solution Repeat steps 1 and 4 until an optimum solution has been obtained

Moving towards optimality Step 4 Modified distribution method MODI / method / U-V method

Step 5

Illustrate Assignment Problem


1. Only one directed arc may be assigned to a specific k, thus Determine Xij0 (i, j=1n) Minimize

Subject to the constraints 2. Only one other city may be reached from a specific city I,

Mathematical formulation of the problem

j=1, 2, 3,.n

and With xij=0 or 1

3. Only one other city can initiate a direct arc to a specified city j, thus

Assignment Problem (AP)


Hungarian method algorithm Routing Problem Maximisation in AP n is based on the concept of opportunity cost and is more efficient in solving assignment problems

4.

Given the kth directed arc ends at city j, (k+1)th directed arc must start at same city j, thus Travelling Salesman problem Forms Constraints of the problem Routing problem

Infeasible Assignment

Unbalanced AP

Illustrate Integer Programming Problem


Optimum solution to an IPP is obtained by using the simplex method ignoring the restriction of integral values The IPP is modified by inserting a new constraint called Gomorys constraint

Gomorys ALLIPP Method Pure Mixed Zero-one

Classification

Integer Programming Problems


Iterative Procedure of IPP

Branch and Bound Algorithm

Step 0: If the master list is not empty, choose an LPP Step 1: Obtain the optimum solution Step 2: If the obtained optimum solution satisfies the integer constraints then record it Step 3: Select any variable that does not have an integer value in the obtained optimum solution to the LPP, chosen in step 0. Add two LPPs to the master list

Step Step Step Step Step Step Step Step

1: 2: 3: 4: 5: 6: 7: 8:

Convert minimisation IPP to maximisation Introduce slack or surplus variables Test integrality of optimum solution Examine constraint equation Express negative fractions Find Gomorian constraint Start a new equation set Arrive at new optimum solution

Illustrate the Infinite Queuing Models


It is a collection of mathematical models of various systems Based on probability concepts Gives an indication of the capacity of a given system Queuing Theory Queue length Number of customers in system Waiting time in queue Waiting time in system Service facility utilisation Arrival pattern Service facilities Queue discipline

Constituents of queuing system Operating characteristics of a queuing system Characteristics of arrival pattern Parameters of service facility

Size of the population Unlimited Limited Patterns of arrivals Behaviour of arrivals

Infinite Queuing Models


Queue Discipline Characteristics of server behaviour

Availability of service Number of service centers Duration of service

Types of Queuing disciplines Characteristics of customer behaviour

The arrangement of service facilities The distribution of service times Servers behaviour Management policies

Dynamic Queuing Discipline Service in Random Order [SIRO] Priority service Pre-emptive Priority

Balking Collusion Reneging Jockeying

Static Queuing Discipline First In First Out [FIFO] Last In First Out [LIFO]

Illustrate the Mathematical Analysis of Queuing Theory


The state, in which the probability distribution remains the same, is called the steady state and the system is said to have acquired a state of statistical equilibrium Statistical Equilibrium Arrivals Service Queuing discipline Z = Average number of arrivals per time P = Traffic intensity n = Number of customers in system c = Number of service channels m = Number of customers in queue E(m) = Average length of queue E(n) = Average number of customers in the system W = Steady waiting time Service System

Properties of Queuing system Notations

Speed of the time Service rate Service time

Mathematical Analysis of Queuing Theory

Configuration of the service system

Limitations

Applications

Models

Classic queuing is too mathematically restrictive to be able to model all real world situations The underlying assumptions of the theory do not always hold well in the real world

Traffic control Determining the sequence of computer operations Predicting computer performance Health services Airport traffic, airline ticket sales Layout of manufacturing systems Telecommunications

Single server-single queue Single server-several queues Several servers- Single queue Several servers- several queues Service facilities in a series Single Channel Models Multiple Service Channels Erlang Family of Distribution of Service Times

Illustrate the Finite Queuing Models


Average number in the queue Average time spent in the queue Statistical distribution of those numbers Probability of the queue being empty Probability of finding the system in a particular state FCFS (First Come First Served) Priority Random LCFS (Last Come First Served) Notations used N = Population (machines, customers etc) M = Service channels (repair men, telephone lines etc) W = Average waiting time H = Average number of units being served L = Average number of units waiting for services F = Efficiency factor X = Service factor

Determines

Queuing Discipline

Finite Queuing Tables

Finite Queuing Models

Steps for Finite Queuing Tables

Measures of System Efficiency

The efficiency factor F is defined as 1. 2. 3. 4. 5. 6. Find mean service and running time Compute the service factor Select the corresponding table Locate service factor value Read values for number of service crews Calculate other measures

Where W is the average time for which a machine has to wait

Illustrate the Simulation


It is an appropriate tool to use in solving a problem It is desirable tool for solving a business problem It provides a trial-and-error movement towards the optimal solution 1. 2. 3. 4. 5. Identify and clearly define the problem List the statement of objectives of the problem Formulate the variables that influence the situation Obtain a consistent set of values for the variables Use the sample obtained in step 4 to calculate the value of the decision criterion 6. Repeat steps 4 and 5 until a sufficient number of samples are available 7. Tabulate the various values of the decision criterion

Reasons for applying simulation technique to OR problems

Methodology Basic concept

Simulation
Sample size Use of Random Number Tables

Simulation is called experimentation in the management laboratory. While dealing with business problems simulation is often referred to as Monte Carlo Analysis

Procedure

Number of trials in a simulation

It is convenient to use a table of random numbers prepared on the basis of some physical phenomenon The grouping of random numbers in the table has no significance Depending upon the number of digits required the random numbers will be chosen in sets of single or two digit numbers

The first step involves coding the data that is you assign random numbers to the variables Then you identify the relationship between the variables and run the simulation to get the results

Illustrate the Simulation Monte-Carlo Method


It is a simulation technique in which statistical distribution functions are created by using a series of random numbers. 1. 2. 3. 4. 5. Define the problem Construct an appropriate model Prepare the model for experimentation Using steps 1 to 3, experiment with the model Summarise and examine the results obtained in step 4 6. Evaluate the result of the simulation 7. Formulate proposals for advice the management Procedure Applications In simulation of probabilistic activity times in PERT network The critical path and the project duration can be found out with a randomly selected activity times for each activity

Definition

Simulation in networks

Simulation Monte-Carlo Method


Advantages The study of very complicated systems or sub-systems can be done with help of simulation We can investigate the consequences for a system of possible changes in parameters in terms of the model The knowledge of a system obtained in designing and conducting the simulation is very valuable It enables us to assess the possible risks involved in a new policy

Simulation in job sequencing

Limitations

It generates a way of evaluating solutions but it does not generate the solution techniques This model does not produce answers by itself Not all situations can be evaluated using simulation It is a time consuming exercise

Illustrate the Project Scheduling and PERT - CPM


Pert was developed in connection with an research and development (R and D) work It is an event-oriented network as in the analysis of a network PERT is normally used for projects involving activities of non-repetitive nature in which time estimates are uncertain CPM was developed in connection with a construction project It is suitable for establishing a tradeoff for optimum balancing between schedule time and cost of the project It is used for projects involving activities of repetitive nature Activities Identify tasks Determine resource requirements Estimate cost and time Specify inner-relationship between tasks Develop network diagram

Planning PERT CPM Phases of PERT and CPM Scheduling Steps Estimate the duration of activities Prepare a time chart showing the start and finish times for each activity use time chart to Calculate the project duration Identify the critical path Carry out resource smoothing exercises

Project Scheduling and PERT-CPM


Probability and Cost Considerations in Project Scheduling Probability considerations are in corporate in project scheduling by assuming the time estimate for each activity is based on three time estimates. They are a = Optimistic time b = Pessimistic time m = Most likely time

Controlling Precedence Relationship

Critical path calculations Determination of the critical path Determination of floats

PERT/CPM Network Components

Events Activities

It refers to comparing the actual progress against the estimated schedule

Potrebbero piacerti anche