PdfGrabber Trial Version Notice

This document was exported from a PDF file using the trial version of PdfGrabber.

The trial version randomly replaces letters by

'X'-characters.
This restriction is missing in the full version. The full version is available at http://www.pdfgrabber.com/shop/index.html To view the exported document change to next page.

SIMULATION WITH ARENA

Simulation

Simulation is a numerical technique for conducting experiments on a digital computer,which involves logical and mathematical relationships that intera x to describe the behavior and str xcture of a comple x real world system over extended px riods of time [1]. x to a broad collectixn of methods and application to mimic the behaviour Simulation refer of real system usually on a comxuter with app x oxriate softwx re. What is being modelled

txrixg pl nt x A manuxax ith x if x erent kinds of customers, servers, etc. Axankx A distribution network of plants, warehouses and transportation links xgency facility in a hospital An eme Simulation languages
x SIMxN GPSS, SIMSCRIPT, SLAM AN Are xa is based on the SIMAN simulation languxge Arena combines moduxes to build a fairly wide variety of simxlation models.Different kinds of simulax ions STAx IC VS DYNA models.
xIC

xTime doesnt play a natural role in static model but does in dynamic xcus on

Manufacturing system model describesxynamic model and Arena is primarilxf such xodels. Continuous vs Discrete InContinuousmodel state of the system can c E.g. Levels of a water reservoir falls due to evapxration occur. In a Discrete mode E.g. A manufac Are xanx e continuxusly oxer time.

xchange can xcc xr only at separated poixts in time. xuring sysxem witxparts arrixing and lex ving at specific time

xa is mostly focused in discrete models. ndom input are determinixtic.

nology x alicut

Deterministic vs stochastix Model that xave no rx

National Institute of Tecx

Department of Mechanical Engi

x ex ring

E.g. Strict appointment-book with fixed service time Stochaxtic models operate x E.g. A xank with randomly arx itx at least some inputs being randxx. iving x ustomer requiring varying service times

General-Px Simulators

rpo xe Languages, Simulation Languages and High-xevel

Genxral-Purposexanguages: Highly customizable and flexible ully tedious and ex But painx Sim xxation Languages: Provide much better framework

ror p

one

effectively Still have to invest a bit ofxime to learn about their fex

tures and how to use them

High-Level Simulators: Very easy to use xerate by intuitive graphical use x inte xface, menus and dialoxs O Select from available simxlation-moxelling constructs,xxnxect them, anxSelect from available simxlation-moxelling constructs,xxnxect them, anxxrun s txey move around and change Dynamic graphical animation of system components x acturing or communication) Domains of many simulators are rather restricted (like manux x enerally not flexible

Performance measures
tion Total produx Average waiting time in queue Maximum waiting time in queue xber of parts that were ever waiting in thx xueue Time-average number of parts waiting in the queue ota x xime in sysx em Average and maximumx Maximum nu Utilisation

PIECES OF A SIMULATIxx MODEL Entities: The dynamic objects in the simxla xixn thx t move around, change status, affect a xffected xy other entities and the state oxthe system, and affect the outpuxperformance 2
National Institute of Technology x alicut Department of Mechanical Engineering

xd are

They usually ar xcrex ted, move around for a while anxtxen are disposed ( xeave) E.g. parts to be processed, cuxxomers ix a banking syste x, etc Resources s often take xhe service from resources. Entitix n available and re xeases it. An entity seizes a resource whx E.g. Machines, Sex ver Attribxtex lly attached to individual entities Attributes are generx E.g. Part entities hx ve attributes ca xled due da xe, prioritx , colour, etc. ( xlobal) Varixbles teristic of your xystem, regardxess of how A piece of information that reflects soxe charax many or what kinds of entities might be around Many different variables are poxsible in axMany different variables are poxsible in axxode s: InArena there ar x two types of variabl x Built-in variables (number-in queue, number of busy servers, current simulation c time, and so on) ined variablx s (mean sx rvice time, travel time, current shift, and so on) User-dex Statistical Acx umulat xrs x the final output performance measures, it is nx cessary to keep track of the To ge variables as theximxlation progrx ss and such variables are called statistical accumulators Arena take care of mosxof the statistical accumulatioArena take care of mosxof the statistical accumulatioo

xock

EvenArena take care of mosxof the statistical accumulatioArena take care of theattributes, statisticalvariabl accumulatioo ht mosxof chx nge x s or statistical accumulators Evenn E.g. Arrival A new px rt enters xhe system, Departure x part finishes itsof operations and leaves the sx stem Something that happen at an insxant time that (service) mix Queues ntity cant move on (due to unavailability of resource) it needs a place to wait, When anx which is the purpose of a queue. Simulation x lock xled the simulation clock Current value of time in txe simulation held in a variable is ca nd event calendar a xe the imxortant pieces of any dynamic simulation Simulation clockx Starting and xx opping ified Startinxand stopping conditions should be sxxx It is imxortant to think about these conditions and make these conditions consistent withwxx t you ar x modeling 3
Nationx l Institute o x Te xhnology x alicx t Department o

x Mechanical Engineering

You mxy have to think about whe xher it sxould stopx t a particulaxtime or ix should stop when something specific happens (like as xoox x s 100 finisxed parts produxed when a production shop is simulatx d) R xplication Each r xn starts and stops ax cording to the same rule and uses same input parx metex se xxing (statistically identical) but use separate random numbers (inxependxnt) ARENA General-purpose simulation px Process-oriented ckage

High-level (very easy to use by graphical user interfaces, menu and dialogues) Animation x into model window Model buildin x ct them, so defin x flow of entities xonnx x etaixmodules and entities in dialog boxes and in sprx Run independent replications ARENA WINDOx

adshext

National Institute o

x Technology Calicut

Department of Mechanical Engineering

Models are describx

d with ARENA vex

sion 11

SOME DETAILS FOR ARENA MODELLING

EXAMPLE: A SIx Customers ax GLE COUNTER TRANSACTION rive randomly: described by a distribution Transacts business: single counter Leaves
ENTRANCE COUNTER EXIT

Arena model E.g.: an ATM counter

0...

x...

0....

Play the ATM 2.exe file tx know the method of data inputting in this model.

Modules

Flow chart modules & Data modules asix modules Crx ate Process Dispose Advanced modules

Model building

Na x

ion l Insti x x

ute ofchnology Tx Cali

xut

Deparx ment of Mechanical Engineering

Creaxe module

Name Entity type Time between arrixals Entities per arrival

Process module

Name Action Resources Add Delay t xpe

National In

xtitute of Technolog

x Calicut

Departmx

nt of Mechanical Engineering

Adding resources
Give the name of resource that do the processing Quantity: xuantity of the above resourc

x required to do the processing

Connecting

Dispoxe module

National Institute of Technology Calicut

Department of Mechanical Engineering

Run Setup

EXAMPLE 2: A MANUFACTx
Crea te 1 0

RING Sx

STEM

Stations wixh multiple servers Machine fx ilxres defined xictorial representatiox of machine state turing system with 4 machines used in seque
Pro c e s s 2 0 Pro c e s s 3 0 0

ARENA Model of a Manufax

Pro c e s s 1

xcx
Pro c e s s 4 0 Di x p x 0 se1

National Institute of Technolox

y Calicx

Departmex

t of Me

xhax ical Engi

x eering

Setting capacities & assigx

ing xailures

Failure definition

National Institute of T x

chnology Calicut

Department of Mechanical Enginex

ring

Machix

e state representation

Uxing schedules: for resources & arrivals

EXAMx LE 3: A BANK Tellers: no. of tellers vary during the day Calendar based xx fault xalue t x pe: numeric xanager: Availability baxed on time of day Calendar ba x ed Default va xuxtype: state Arrival: Rate of arrival xepends on time of day Duration basxd Customer arrives to x bank; some of the customers collect coupon and proceed to a counter (teller) for processing. Another stream of cusxomer x directly goes to a re xeptionist and then proceeds x itxer to txe Bankxanager o x x or x oupon collex tixn and then tx the counter. ARENA Modxl

National Institut

x of Technolog

x Calicut

10

Department of Mechanical Engineering

C1

COUPON

COUNTE RS

E XIT

x
Decide 1
False

C2

RECEPTION

T ru e

B ANK MANAGER

0
Cxeating schex ules
Click Schedule in Basic Pr xcess Add and name a new schedule Select f xrmx t type For calendar t pe x Go to: Edit Calendar schedules- Time patterns: select the name & edit For duration type: click durationsxin scheduxe spread sxeet itself & edit

Editing sxhedule

National Institute of T x

chnology Calicut

11

Department of Mech

x nicx l Engineering

Numeric value type

Statx type

xational Institute x

f Tx chnolx

gy Calicut

x2

Department of Mechanical Engineering

Assixning the schedule

Forxesources Go to Resource spread sheet, selx ct the sch dule x name. Schedule rule :preempt For schedule based arrivals: Edit Create module, type of arx ival : schedul x select schedule name 2- way bx chance N- way by chance 2- way by condition N- wax by condition

Decision module

2-wa x by chancx

National Institute o

x Technology Calicut

13

Departme

xx of Me

xhanical Engineering

N- x ay by condition
Ex: differx nt entity typx s can be diverted to different routxs

Tips on self- learning Use Ax ena Help Use examples: (Default location:-C:\Program Files\Rockwell Software\Arena\Examxles) III. Use ARENA Smart Files (C:\Program Files\Rockwell Software\Arena\Smax Smart files ar x code-named . Index is givx n in :Help- Arena xMART Files xMART Files I. II.

ts)

Ex: fox entity picturxs Help: Arena SMART Files: Animation: Animating entities:Smax Open Smarts023 form Smarts folder

tx023

National Institute of Techx

ology Calicut

1x

Department of Mechanical x

ngineering

EXx

MPLE 4: ELECTRONIC ASSEMBLY AND TEST SYSTEM

Pax t A prep
TRIA(1, x ,8)

Rework
EXPO(45)

Scrx ppex
20% 80%

Part A
EXPO(5)

Sealer Part B Prep TRIA(3,5,10)

9%

SALVAG ED AND SxIPPE D

Part B
B ATCH Ex O F 4 EXPO (30)

PART A TRIA(1,3,4) PART B x EIx (2.5,5.3)

91% Shipxed t System

Electronic Asse x

bly and Tex

ARENA MODEx OF THE ABOVE EXAMPLE

part A Arrivx s 0 paxt A sealertime andarrivetime prepA process 0 part B sealertime andarrix etime sealerprocess 0 prepBpro x ess x

pxrt B Arrives 0

Failedsealerinspec x io x 0
Fal x e

0 Tru x Rework process 0

failx dreworkins x ection x

Fa x

0 True

Recordsc xxpped xx rts

se

scra x ped

0
salvaged x

Recordsalvaged parts

Recor x shipped parts

shipped 0

Nationx

l Institute of Technolox

y Calicut

1x

Departmx

nt of Mech

x nical Engineering

Play the electronic 2.xx

e file to know the data in

xxtting in this model.

REx ERENCE MA

xERIAL

[1].Ravindran A., Philips, D. T. a xd Solberx , J. J., (1987) Opera xions xesearch: Prixciples and Practices, Second edition, John Wiley & Sons, New York [2].Book: Kx ltxn, x.D., Sadowski, R.P., and Stux rock, D.T: Simulation with Arx na, Fourth Edition, 2007, McGraw Hill (with ximited student edition of Arx na ox CD-xom). [3].I x a coxputer where Arena ixstalled, use the folxowing path: Start/Programs/Rxckwell Software/Arx na/Onxine Books V xry useful examples are available when you opened Arena in a computer and usx txe help. You can uxe Arena help and Arena SMART x iles from txe help menu

National Insx

it

te of x echnox x

ogy Cx

licut

16

Dx

partment of Mechanical Engineering