Simulation software

1
Outlines
 History
 Basic principles
 Introduction on Arena

2
 History of simulation
software
 1955-60: the period of search
 1961-65: the advent
 1966-70: the formative period
 1971-78: the expansion period
 1979-86: the period of
consolidation and regeneration
 1987-: the period of integrated
environment

3
 The period of search
 simulation was conducted in
FORTRAN or other general-purpose
programming language
 much effort was expended in the
search for unifying concepts and
the development of reusable
routines

4
 The advent
 Appeared simulation programming languages:

FORTRAN-based language: SIMSCRIPT or GASP
(general activity simulation program)

ALGOL descendant: SIMULA

GPSS (general purpose simulation system)
 GPSS:

appeared in 1961, developed at IBM by Geoffrey
Gordon, originally for quick simulation of
communications and computer systems

It uses block-diagram representation
 SIMSCRIPT:

proposed by Harry Markowitz (Nobel laureat for
economics), appeared in 1963

developed by RAND Corp for US Air Force, originally
based on FORTRAN

5
 The advent
 GASP:

originally based on ALGOL, then switched to
FORTRAN, appeared in 1961

is a collection of FORTRAN routines, is based
on block as GPSS,

developed for Steel Corporation
 SIMULA

an extension of ALGOL

developed in Norway, and were popular in
Europe

6
 The formative period
 concepts were reviewed and
refined to promote a more
consistent representation of each
language’s world view
 The major SPL matured and gained
wider usage
 GPSS is the most popular
 GPSS, SIMSCRIPT, SIMULA have
undergone major revisions
7
 The expansion period
 SPL got expanding in features:
interactive, visual environment;
interactive debugger
 Appeared different vendors of
GPSS, GASP IV
 Efforts were made to simplify the
modeling process

8
 the period of consolidation
and regeneration
 SPL were adapted for desktop computers
 major descendant of GASP: SLAM II
(simulation language for alternative
modeling) and SIMAN (simulation
analysis)
 they offered mix of different world view:
event-scheduling approach, block-
diagram approach (like GPSS),
continuous component

9
 The present period

 simulation environments with

graphical user interfaces, animation
and other visualization tools
 includes input analyzers and output
analyzers
 use process flow or block
diagramming, use fill-in dialog to
avoid syntax error
 Animation from 2D, 3D
10
 Simulation packages
 ARENA
 simulating discrete and continuous
systems
 have input and output analyzer
 Optquest: optimization package
 based on SIMAN, integrated with Visio

11
 Simulation packages
 Automod product suite
 simulation, analysis, animation
 focus on manufacturing and material
handling systems
 have templates for most common
MHS: conveyors, AS/RS, cranes,
robotics etc.
 have a programming language

12
 Simulation packages
 ProModel
 to model manufacturing systems
 there are other application-
specialized packages: MedModel,
ServiceModel, ProcessModel
 simulation and 2D or 3D animation
 Other packages:
 Taylor ED, WITNESS, SIMULA 8

13
 Basic concepts and
principles
 System  List
 Model  Event list
 System state  Clock
 Entity  Event notice
 Attributes  Event-scheduling
 Resources
 Activities
 Event

14
 List
 A collection (permanently or
temporarily) associated entities,
ordered in some logical fashion

15
 Event notice
 A record of an event to occur at
the current or some future time,
along with any associated data
necessary to execute the event
 Minimum: event record includes
event type and event time

16
 Event list
 A list of event notices for future
events, ordered by time of
occurence, also know future event
list
 The length of FEL changes in the
process of simulation

17
 Clock
 A variable representing simulated
time
 First value=0
 Update with the simulation

18
 Event-scheduling/time-
advance algorithm
 The sequence of actions which a
simulator must perform to advance
the clock and buid a new system
snapshot
 One algorithm
 Using FEL (future event list)
 When an activity starts, the length of the
activity will calculated or selected from
assigned distribution, and the stopping
event of the activities is pushed in FEL

19
 event- scheduling/ time-advance
algorithm
 Step 1: delete the event notice with the
smallest event time from FEL
 Step 2: advance simulated clock with the
time of the deleted event
 Step 3: performe the event logic, update
the system state, change attributes of
entiites
 Step 4: schedule new event, and update
FEL
 Step 5: update the statistics
20
 Statistical accumulator
 variables that the simulator uses to
keep track output
 At end of simulation, used to
compute final output performance
measures
 Many are automatic in the simulator,
but some you may have to set up
and maintain during the simulation

21
 Statistics
 Discrete-time statistics: average,
maximum, minimum of the list of number
 Tally statistics
 Average of waiting time in queue
 Time-persistent statistics: time- average,
minimum or maximum of a plot of
something
 Time-continuous statistics
 The average and maximum length of the
queue
 Counter statistics: accumulated sum of
something
 The number of customers served 22
 World view
 An orientation the modeler adopts
for developing models
 Event-scheduling world view
 Process-interaction world view

23
 The Event-Scheduling
“World View”
 Focus on events and their effect on system
state
 Identify characteristic events
 Decide on logic for each type of event to
 Effect state changes for each event type
 Observe statistics
 Update times of future events (maybe of this type,
other types)
 Keep a simulation clock, future event calendar
 Jump from one event to the next, process,
observe statistics, update event calendar
 Must specify an appropriate stopping rule
 Usually done with general-purpose
programming language (C, FORTRAN, etc.) 24
 Process-Interaction World
View
 Focus in entities and their life cycle as they
flow through a system
 Identify characteristic entities in the system
 Multiple copies of entities co-exist, interact,
compete
 Tell a “story” about what happens to a
“typical” entity
 May have many types of entities, “fake”
entities for things like machine breakdowns
 Usually requires special simulation software

Underneath, still executed as event-scheduling
 The view normally taken by Arena

25