MODELING AND SIMULATION -AN INTRODUCTION

By, Dr.N.SELVARAJ Asst.Professor Dept. of Mechanical Engg. NIT ,Warangal.

Overview
What is simulation ? When simulation is appropriate tool When simulation is not appropriate Advantages and Disadvantages of simulation Areas of application Systems and Systems environment Components of a system Discrete and continuous systems Model of a system Types of models Discrete-event systems simulation Steps in simulation study Simulation Software's and selection of software's Literatures

What is model?
A model is a representation of real system or process for the purpose of studying the system A model may incorporate logical, mathematical and structural aspects of the system or process

What is model?
Model set of assumptions/approximations about how the system works
Study the model instead of the real system usually much easier, faster, cheaper, safer Can try wide-ranging ideas with the model

Make your mistakes on the computer where they dont count, rather than for real where they do count

Often, just building the model is instructive regardless of results

What is Simulation?
Simulation very broad term methods and applications to imitate or mimic real systems, usually via computer Applies in many fields and industries Very popular and powerful method

What is Simulation ?
The representation of real situation in a computer by means of model so that, different conditions can be tested over period of time. It is a particular type of mathematical model of the system or representation (Imitation) of real world system or process over period of time.

System vs. Its Model

Simplification
Real System

Abstraction Assumptions

Model

Model of a System
A model is defined as a representation of a system for the purpose of studying the system. Types of models : Mathematical or physical. A mathematical model uses symbolic notation and mathematical equations to represent a system. A simulation model is a particular type of mathematical model of a system.

Model Classification
Physical (prototypes) Analytical (mathematical) Computer (Monte Carlo Simulation) Descriptive (performance analysis) Prescriptive (optimization)

Discrete and Continuous Systems

Discrete systems: State variables change only at a discrete set of points in time.

Example : bank: the number of customers change when a customer enters or leaves the system

Continuous system: State variables change continuously over time. Example: water level in the dam and body temperature

3
2 1 time t

water

time

Physical (Prototypes)

Analytical (Mathematical)

n P( N n) 1

Single Stage Queuing Model

1 W L W

Wq W

2 Lq Wq

Descriptive
(Performance analysis)

Simulation vs. Real World

Simulation Models
Simulation models: Static or dynamic, Deterministic or stochastic, Discrete or continuous. Static or Monte Carlo simulation represents a system at a particular point in time. Dynamic models: represent systems as they change over time. Deterministic simulation: Known sets of inputs and a unique sets of outputs.

If all patients arrive at an appointed time contd

Simulation Models
Stochastic simulation: has one or more random variables as inputs. Random inputs lead to random outputs.

Bank: random inter-arrival times and random service times. Output: average number of people waiting, average waiting time of a customer.

Discrete-event simulation: Modeling of systems in which the state variables changes at a discrete set of points in time .

Steps in Simulation study

Problem formulation:

Statement of the problem. The problem should be clearly understood. The objectives indicate the questions to be answered by the simulation. Determination should be made whether the simulation is appropriate methodology or not.

Setting of objectives and overall project plan:

Steps in Simulation study

Model conceptualization:

It is an art. Abstract the essential features of the problem Select and modify basic assumptions that characterize the system Enrich and elaborate the model until a useful approximation results. It is best to start a simple model and build toward greater complexity. Not necessary to have one-one mapping with real system. Only the essence of the real system is needed. Advisable to involve model user However, only experience can teach the model building.

contd..

Steps in Simulation study

Data collection:

Different kinds of data should be identified and collected while building a model.

Model translation: The model can be translated into program. Verified:

Is the computer program performing correctly ? With complex models it is difficult.

Validation: determination that a model is an accurate representation of the real system

contd..

Experimental design:
Length of the initialization period The length of simulation runs The number of replications to be made for each run Documentation and Reporting

Two types of documentation: program and progress. Reporting frequent deliverables.

Implementation: Completion of previous steps

PERFORMANCE TO BE MEASURES
Throughput Cycle time Resources utilizations WIP Staff requirements Queueing delays Bottleneck Effectiveness of scheduling Effectiveness of control system

When Simulation is appropriate tool

Simulation enables the study of internal reactions of a complex system Informational, organizational, and environmental changes can be simulated. To improve the system performance Important variables that affect the system can be identified To experiment with new designs or policies prior to implementation, so as to prepare for what may happen. The modern system is so complex that the interactions can be treated only through simulation.

When simulation is not appropriate

If the problem can be solved with common sense.

Average arrival rate 100/hour and service rate is 12/hour, then the number of servers 100/12=8.33. Which means 9 or more servers are needed.

If the problem can be solved analytically. If it is easier to perform experiments If costs exceed savings If the resources or time is not available. If data or estimates are not available Ability to verify the model If managers have unreasonable expectations If the system is too complex.

Advantages
New policies, operating procedures, decision rules, information flows, or organizational procedures, and so on can be explored without disrupting ongoing operations of the real system New hardware designs, physical layouts, transportation systems, and so on, without committing resources for their acquisition. Hypotheses about how or why certain phenomena occur can be tested for feasibility. Insight can be obtained about the interaction of variables contd..

Advantages
Insight can be obtained about the importance of variables to the performance of the system Bottleneck analysis can be performed where work in progress , information, materials and so on are being delayed. A simulation study can help understanding how the system operates rather than how individuals think the system operates. What-if questions can be asked. This is useful in the design of new systems.

Areas of Applications
Manufacturing applications Construction Engineering Military applications Logistics, transportation, and distributed applications Business process simulation Human Systems

Manufacturing Applications
Analysis of electronics assembly operations Design and evaluation of a selective assembly static for high-precision scroll compressor bells Comparison of dispatching rules for semiconductor manufacturing using large-facility models Evaluation of cluster tool throughput for thin-film head production Determining optimal lot size for a semiconductor back-end factory
contd..

Manufacturing Applications
Optimization of cycle timer and utilization in semiconductor test manufacturing Analysis of storage and retrieval strategies in a warehouse Investigation of dynamics in a service oriented supply chain Model for an Army chemical munitions disposal facility

Human Systems
Modeling human performance in complex systems Studying the human element in air traffic control.

Systems and System Environment

A System is defined as a group of objects that are joined together in some regular interaction or interdependence toward the accomplishment of common purpose

Example: Production system manufacturing automobiles. The machines, component parts and workers operate jointly along an assembly line to produce a high-quality vehicle.

Environment: A system effected by changes occurring outside the system. Such changes are said to occur in the system environment. There is a boundary between the system and environment.

Components of a System
Entity: object of interests in the system Attribute: property of an entity Activity: time period of specified length

Example: bank: Customers might be one of the entities, balance might be an attributes, and making deposit is an activity.

contd..

Components of a System
State: Collation of variables necessary to describe the system at any time relative to the objectives of the study.

Bank: # of busy tellers, # of customers waiting in the queue, arrival time of the next customer

Event: It is defined as an instantaneous occurrence that may change the state of the system. Depending on purpose, the number of components (entities, attributes, activities, states, events) varies.

Example of Systems and Components

System
Banking

Entities
Custome rs

Attribu tes
Checking Account balance

Activiti es
Making deposits Machini ng, welding Transmit ting Withdra wing

Events
Arrival and departure

State Variable
Number of busy tellers, number of customers waiting

Production

Machine s

Speed, capacity, breakdo wn rate Length, destinati on Capacity

Break down

Status of machine (busy, idle, down) Number waiting to be transmitted

Levels of inventory

Communic -ation

Arrival at
destination

Messages
Ware house

Inventory

Demand

Modelling Approach
Event Scheduling

System modelled via characteristic events Events have subroutines which update state variables. Time oriented sequence of inter-related events that describes the experience of an entity as it flows through a system. Overlay to an event scheduling system. Approach adopted in most current software.

Process Orientation

SIMULATION SOFTWARE

1st Category

2nd Category

3rd Category

Channel purpose language

FORTRANC, C + +VB, VB+ + . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .. . . . . . . . . . .many other oriented languages

Simulation language
GPSS (1965)SIMSCRIPT (1963)SIMULA GASP (1961)ALGOL SLAM (1979)SIMAN GPSS/4 (1977)SLAM IIAWESIM (1995)GEMS

Simulation Packages
ARENA (1993)AutoMODQ UEST EXTEND PROMODEL TaylorED WITNESS. . . . . . . . . . .and many more

Webbased simulation

JAVASIMWEBBASED SIMULATION. . . . . . . . . . .. . . . . . . . . . .. . . . . . . . . . .. . . . . . . ....

Table 2 : Commercial Simulation Language

40

Simulation Softwares
General purpose simulation packages: Arena,Extend, Awesim, Symix, GPSS, Micro saint, Modsim, Simple+, Taylor etc..

Simulation Software's
Application Oriented simulation packages Manufacturing: Automod, Promodel, Quest, Flexsim, Extend, Auto shed, Taylor, etc.. Communications Networks: COMNET III, OPNET Modeler, IT Decision guru etc.. Process Reengineering: Process model, SIMPROCESS, Extend BPR etc..

Simulation Languages
General in nature
Can model almost any type of system. Frequently include specific modelling constructs (such as material handling systems). Steep learning curve. Significant modelling and programming expertise is necessary. Long(ish) development cycles.

Common Features
1. 2. 3.

4.
5. 6. 7.

Generating random numbers (i.e. ~U(0,1)) Generating random variates from a specified probability distribution. Advancing the simulation clock. Determining the next event on the list event and passing control of to the appropriate piece of code. Adding and deleting records from a list. Collecting output statistics and reporting the results of the simulation run. Trapping error conditions.

GPSS Code
Assume an M/M/1system. Interarrival time = 2.0 minutes (exponential) Service time = 1.0 minutes (exponential). Assume an infinite queue.
* Simulation of M/M/1 system SIMULATE GENERATE RVEXPO(1, 2.0) QUEUE SERVQ SEIZE SERVER ADVANCE RVEXPO(2, 1.0) RELEASE SERVER TERMINATE 1 CONTROL STATEMENTS START END 1000

* * *

SLAM
IBM stopped support and development of GPSS about 1972. A market developed for alternative software that could run on newer machines (VAX & UNIX). In 1979, Alan Pritsker and David Pegden create SLAM (Simulation Language for Alternative Modeling). In the early 80s Pritsker and Pegden develop SLAMII, which ran on engineering workstations.

A feature of this new language is a graphical model builder. Users enter their model as a network diagram. When complete, the network is translated into SLAM code.

SLAM-II Code or AWESIM

Originally one of the slowest components of a SLAM model was compiling. Compiling really translates the model into a set of FORTRAN subroutines. To speed up compiles, controls were separated from the main body of the model. Controls were designed to be short and changeable, while models were to be big and relatively fixed.
1 2 3 4 5 6 7 8 ; 1 2 3 4 5 GEN; LIMITS; INITIALIZE,0.0,1000,YES,,NO; NET; FIN; RESOURCE,,SERVER,1,{1}; CREATE,EXPON(2,1),0.0,,INF,1; ACTIVITY; AWAIT,1,{{SERVER,1}},ALL,,NONE,1; ACTIVITY,1,EXPON(1,1); FREE,{{SERVER,1}},1; ACTIVITY; TERMINATE,INF;

Awesim

SIMAN
About 1983 or so, Dennis Pegden develops his own simulation language. SIMAN (SIMulation ANalysis). The language was designed to run on a PC. It is remarkably similar in look, feel, content and style to SLAM. A lawsuit entailed.
BEGIN;

CREATE,,EX(2,1); QUEUE, 1; SEIZE: SERVER; DELAY: EX(1,1); RELEASE:SERVER:DISPOSE;

END; BEGIN; DISCRETE, 1000, 1, 1; RESOURCES: 1, SERVER; REPLICATE,1; END;

SIMAN AND SLAM

SIMAN is tailored for the PC market. SLAM remains focused on workstations. SIMAN introduces an animation package (CINEMA) about 1985 or so.

The animation is an add on unit for the model. Originally it required specialized (& expensive) hardware.

SLAM responds with a PC version of SLAM in the late 1980s (which also has animation). Both firms develop software to integrate factory scheduling into simulation runs.

WITNESS
WITNESS is offered by the Lanner Group. WITNESS is strongly machine oriented and contains many elements for discretepart manufacturing. WITNESS models are based on template elements. Elements may be combined into a designer element module to be reused

Witness (Lanner Inc)

Simple building block design Interactive Full range of logic and control options Elements for discrete manufacture, process industries, BPR, e-commerce, call centers, health, finance and government Statistical input and reports Link system to other software easily (CAD/Excel) Optional 3D/VR views Model Optimisation

$13,000-$17,000 ($US)

Witness (Lanner Inc)

Witness (Lanner Inc)

Witness (Lanner Inc)

ARENA
Process hierarchy. Integrates with Microsoft desktop tools Spreadsheet interface Crystal reports Free runtime software. Fully graphical environment. No programming required. VBA embedded. Optimization with OptQuest for Arena. Builds reusable modules. $1,000 - $17,000 ($US). Various add-in modules available.

ARENA
Arena can be used for simulating discrete and continuous systems Arena employs an object based design for entirely graphical model development. Modules are organized into collections called templates.

57

ARENA

58

ARENA

59

GPSS/H
Successor to the orginal simulation language (GPSS).

Was freeware on IBM 360s

Makes use of common program blocks. Proven, reliable software. Extremely flexible. Extensive error checking routines. Post-process animations (Proof) can be built. ~$5,000 ($US)

AutoMod
It includes the AutoMod simulation package, AutoStat for experimentation and analysis, and Auto View for making AVI movies of the built-in 3-D animation. The main focus of the AutoMod simulation product is manufacturing and material handling systems.

61

Automod
Combines Virtual Reality (VR) graphics with a discrete and continuous simulation environment.

Manufacturing operations Material handling systems Tanks and pipe networks IC Manufacturing Transportation and logistics systems

$15,000 - $100,000 ($US)

AutoMod

63

AutoMod

64

AutoMod

65

Production flow

QUEST
QUEST if offered by Deneb Robotics QUEST models are based on 3-D CAD geometry. A QUEST model consists of elements from a number of element classes. Built-in element classes include AGV and transporters, buffer, conveyor, labour, machine, parts and process. Each element has associated geometric data and parameters that define its behaviour

QUEST

QUEST

Flexsim (Ware house)

Flexsim

Flexsim

ProModel
ProModel is offered by ProModel Corporation It is a simulation and animation tool designed to model manufacturing systems. ProModel offers 2-D animation with an optional 3-D like perspective view.

ProModel
ProModel is offered by ProModel Corporation ,USA. It is a simulation and animation tool designed to model manufacturing systems. ProModel offers 2-D animation with an optional 3-D like perspective view.

Promodel
State-of-the-art simulation engine Graphical user interface Distribution-fitting. Output analysis module Optional optimizer. Modules designed for: Manufacturing Healthcare Services $17,000 ($US)

(Run Hours 231.57)

Graph - 2 : Case Study - 2

2900 2700
Cycle Time

2500 2300 2100 1900 1700 1500 1 2 3 4 5 6 7 8 Number of Vehicles GA Heuristic

from above table average process time in percentage of total scheduled hours = (39.62+14.97+20.04+11.88+11.16+21.7+16.99+22.89+23.89+20.51)/18 = 11.31%=0.1131
average process time = 0.1131*231.57*60=1572.05

Assembly line (Promodel)

Production flow line

METHODOLOGY FOR SELECTION OF SIMULATION SOFTWARE

Stage 1 Need for purchasing simulation software

Stage 2

Initial software survey

Evaluation

Stage 3

Stage 4

Software selection
Software contract negotiation

Stage 5

Stage 6

Software purchase

Figure 3 : Stages of simulation software selection methodology

Need for purchasing simulation software

Purpose of simulatio n

Constrai nts

Models to be simulate d

Model develope rs

Educati on

Quick and dry ind

D/C ind or researc h

Time

Discret e

Contin uo.

Combi ned disc/co nt

Individ ual prefere nce

Previous exper. in simulation

Initial software survey

Continued in the next slide

Short list of software for evaluation

Initial software software survey

Initial

survey

Initial software survey

Initial software survey

Initial software survey

Initial software survey

Results of Evaluation

Software selection

Selection of software Legend:

Software contract negotiation

Stages

Contract acceptable

Intermediate Results

Software purchase

Elements

SELECTION OF SIMULATION SOFTWRAES

Modeling Building features Graphical model building Input and out analysis capability Conditional routing Specialized components and templates Interface with general programming

SELECTION OF SIMULATION SOFTWRAES

Runtime environment Execution speed,Model size Number of variables and attributes Model status and statistics

SELECTION OF SIMULATION SOFTWRAES

Animation and layout features

Type of animation, Dimensions, Movement, Quality of motion, hardware requirements

SELECTION OF SIMULATION SOFTWRAES

Out put features

Graphical, text, animations etc.. Vendor support and Product documentation

Training, documentation, Tutorials, easy to contact, etc..

4.MAJOR SOURCES OF ERRORS IN SIMULATION

Modeling errors
Analysis errors or Programming errors Sampling errors

Modeling Errors

The use of invalid models may results in serious simulation errors, and in fact this the cause for the failure of many simulation errors. A valid model is necessary prerequisite in simulation by using V & V

Analysis errors or Programming errors

Coding error Numerical error Random number error Random variate error

Coding error
Coding error. The code wrong, in commercial software or the practitioners specification of the model. Sometimes called verification error (analogous to modeling error being called validation error).

Numerical error
Numerical error. Computer arithmetic is not real number arithmetic; computers can store only a finite set of numbers. Examples include numbers close to zero being denser than numbers close to one, floating-point comparisons being suspect because of rounding, and combinatorial calculations overflowing.

Random-number error
Random-number error. Pseudorandom numbers are not truly random numbers. As computers become faster, sample sizes become larger, and sensitivity to randomnumber error increases.

Random-variate error
Random-variate error. Methods to generate random variates are sometimes approximations. For example,the reasonably good standard normal inverse transformation x = (u0.135 (1u)0.135)/0.1975truncates at about five standard deviations from the mean.

Sampling errors
Sampling error. Monte Carlo simulation analysis is fundamentally a statisticalinference method;therefore, sampling error is unavoidable. Sampling error is typically measured by standard error, the standard deviation of the point estimator, which is often inversely proportional to the square root of the sample size.

Error control in simulation

Detection and control of bias(warm-up period) Replication of simulation run Batch mean Regenerative method initial

CONCLUSIONS
the application for simulation to address manufacturing problems. Developments in the area of simulation existing softwares for discrete event simulation and conduction of simulation studies were reviewed. The necessity and importance of simulation for modeling and analyzing the various classes of manufacturing problems was focused in this paper; we hope this paper may encourage the extensive use of simulation in manufacturing and development of simulation technology for addressing the problems which need serious attention.

MANUFACTURING SIMULATIONLABORATORY
List of Experiments
Performance evaluation of single line multi stage manufacturing system 2. Performance evaluation of multi line multi stage assembly manufacturing system 3. Break down analysis of manufacturing system 4. Study of Just In Time manufacturing system 5. Job shop scheduling problem 6. Design of Flexible Manufacturing system
1.

All the above problems are studied with different manufacturing environment, like line balancing, unbalancing, different buffer size, stochastic etc.. Using Awesim, Promodel, Automod, Flexsim and Witness simulation softwares.

MANUFACTURING SIMULATION LABORATORY

Hardware: 1. Pentium III (Wipro make) -----------------------1 No 2. Pentium II Server (Compaq make) --------------1 No 3. Intel Celeron Computer(Compaq make)--------9 Nos 4. Intel Celeron Computer(DTK make)------------2 Nos Software: 1. Automod 10.0 ----------- 11 Network user licenses 2. Awesim 3.0--- ----------- 11 Network user licenses 3. Promodel 4.22 ----------- 11 Network user licenses 4. Flexsim 3.0 --------------- 20 Network user licenses 5. Witness 2.1---- ----------- 10 Network user licenses Activities: 1. Lab. Classes for M.Tech (Manufacturing Engg) and M.Tech( CIM) 2. Design Project works for M.Tech and B.Tech 3. Research works

Literatures
www.wintersim.org www.lionhrtpub.com Int.Jour. of Simulation and Modeling Int.Jour. of Statistical computation and simulation Simulation modeling and analysis by Law and Kelton Discrete event system simulation by Jerry banks and Nelson System Simulation by Jordon