Prof M S Prasad
,
Based on open literature and reports . For special system model see Session II presentations.
@2008
Introduction to Simulation
A Simulation is the imitation of the operation of a
real-world process or system over time
A System is defined to be a set of elements which interact or interrelated in some fashion
Elements having no relationship with the set of
elements that have been chosen as system can not affect the system hence irrelevant A System may consist of sub systems or may be a part of a larger system
05.06.2008
Elements that often make up the system are called Entities Entities that comprise a system need not be tangible e.g, a queuing system is made up of customers, queue and servers Customers and servers are physical entities but queue itself is a concept
05.06.2008
More components of a system
An Attribute is a property of a system An Activity represents a time period of specified
length State of system is defined to be that collection of variables necessary to describe the system at any time , relative to the objective of the study
In the study of a bank possible state variables are number of
busy tellers, number of customers waiting in the queue or being served, arrival and service times of the next customer
An Event is defined as an instantaneous occurrence
that may change the state of the system
05.06.2008 4
More terms of a system
Endogenous used to describe the activities and
events occurring within a system Exogenous is used to describe activities and events in the environment that affect the system
In the bank arrival of a customer is exogenous event
and completion of service of a customer is endogenous event
05.06.2008
Examples: Production System
Entities Machines
Attributes (property of an entity) Speed , Capacity,
Breakdown rate Activities (time period of specified length) Welding, Cutting, Stamping Events breakdown State variables Status of machines busy, idle or down
05.06.2008
Examples: Communications System
Entities Messages
Attributes (property of an entity) Length ,
Destination Activities (time period of specified length) Transmitting Events arrival at destination State variables Number of messages waiting to be transmitted
05.06.2008
Examples: Inventory System
Entities Warehouse
Attributes (property of an entity) Capacity Activities (time period of specified length) Issue,
Receipt Events Demand State variables Level of inventory, Backlogged demands
05.06.2008
Introduction to model
A model is a system that is used as a surrogate for another system Reason for using a model
Helps in understanding the behaviour of a real system
before it is built Cost of building and experimenting with a model is less Models can be used to mitigate risk pilots can be taught how to cope with wind sheer while landing Models have the capability of scale time or space in favourable manner wind sheer can be produced on demand
05.06.2008 9
Types of Models
Broadly there are two types
Physical
(Scale models, prototype plants,)
Mathematical
(Analytical queuing models, linear programs, simulation)
05.06.2008 10
Ten Types of Models
Iconic - physical models that are images of the real world; dimensions are usually scaled up or down; for example, models of cars might be constructed and tested in a wind tunnel Analog - model that substitutes one set of properties for another; may be iconic or mathematical; electric resistance often used as an analog of the friction of a fluid flowing in a pipe;
05.06.2008
11
Ten Types of Models Stochastic - probabilistic model that uses
randomness to account for non measurable factors (e.g., weather) Deterministic - model that does not use randomness but uses explicit expressions for relationships Discrete - model where state variables change in steps as opposed to continuously with time (e.g., number of cattle in a barn); may be deterministic or stochastic
05.06.2008 12
Ten Types of Models
Continuous - model whose state variables change
continuously with time (e.g., biomass in a field); usually sets of differential equations used; initial conditions required (can be difficult to obtain for some systems!) Combined - model where some state variables change continuously and others change in steps at event times; for example, a field of hay might be modeled using a combined approach with the biomass modeled continuously during growth and then as a discrete event when harvested
05.06.2008 13
Types of Models
Mathematical - abstract model usually written in
equation form Object-oriented - use objects that are abstractions of real world objects and develop relationships and actions between objects; comes from field of artificial intelligence Heuristic - heuristics (rules) are used to model the system; comes from field of artificial intelligence.
05.06.2008
14
Static Model Dynamic Model Lumped & distributed Models
: distributed model use Partial diff equation to Explain space varying parameters. In Lumped model the space variations are defined in finite numbers making it a differential eqn,
05.06.2008
15
What is Simulation?
A Simulation of a system is the operation of a
model, which is a representation of that system. The model is amenable to manipulation which would be impossible, too expensive, or too impractical to perform on the system which it portrays. The operation of the model can be studied, and, from this, properties concerning the behavior of the actual system can be inferred.
Introduction 16
Applications:
Designing and analyzing manufacturing systems Evaluating H/W and S/W requirements for a
computer system Evaluating a new military weapons system or tactics Determining ordering policies for an inventory system Designing communications systems and message protocols for them
Introduction
17
Applications:(continued)
Designing and operating transportation facilities such
as freeways, airports, subways, or ports Evaluating designs for service organizations such as hospitals, post offices, or fast-food restaurants Analyzing financial or economic systems
Introduction
18
Types of Simulation Models
System model
deterministic
stochastic
static
dynamic
static
Monte Carlo simulation
dynamic
continuous
Continuous simulation
05.06.2008
discrete
Discrete-event simulation
continuous
Continuous simulation
discrete
Discrete-event simulation
19
Stochastic simulation: a simulation that contains random (probabilistic) elements, e.g.,
Examples Inter-arrival time or service time of customers at a restaurant or store Amount of time required to service a customer Output is a random quantity (multiple runs required to
Stochastic vs. Deterministic
analyze output)
Deterministic simulation: a simulation containing no random elements
Examples Simulation of a digital circuit Simulation of a chemical reaction based on differential equations Output is deterministic for a given set of inputs
05.06.2008 20
Static vs. Dynamic Models
Static models Model where time is not a significant variable Examples
Determine the probability of a winning solitaire hand
Static + stochastic = Monte Carlo simulation Statistical sampling to develop approximate solutions to numerical problems
Dynamic models Model focusing on the evolution of the system under investigation over time
05.06.2008
21
Continuous vs. Discrete
Discrete
State of the system is viewed as changing at discrete
points in time: arrival of a customer in a queuing system An event is associated with each state transition
Events contain time stamp
Continuous
State of the system is viewed as changing continuously
across time: rise if water level in a dam System typically described by a set of differential equations
Few systems in practice are wholly discrete or continuous
05.06.2008 22
Discrete & Continuous Systems
Essential to remember that
A discrete simulation model is not always used to model
a discrete system Similarly, a continuous simulation model is not always used for a continuous system Simulation models may also be mixed both discrete and continuous
Choice of discrete or continuous simulation
models is a function of
Characteristics of the system Objective of the study
05.06.2008 23
Discrete & Continuous Systems
Communication channel Modeled as discrete if characteristics of movement of each message is important Modeled as continuous if flow of messages as aggregate over the channel is important In this course we will study only Models that are discrete, dynamic and stochastic
05.06.2008
24
Discrete event systems (DES) DES are dynamic systems which evolve in time by the
occurrence of events at possibly irregular time intervals DES abound in real-world applications
Examples include traffic systems flexible manufacturing systems computer-communications systems production lines flow networks.
05.06.2008
25
Discrete event systems (DES)
Most of these systems can be modeled in terms of
discrete events whose occurrence causes the system to change from one state to another In designing, analyzing and operating such complex systems, one is interested not only in performance evaluation but also in sensitivity analysis and optimization.
05.06.2008
26
Discrete event system simulation (DESS)
It is modeling of systems in which the state variable
changes only at a discrete set of points in time Simulation models are analyzed by numerical methods rather than by analytical methods Analytical methods apply deductive reasoning to solve
Differential calculus can be used to calculate EOQ
In case of simulation model is run rather than
solved An artificial history of the system is generated (with the help of computer) based on system characteristics and observations are collected to be analyzed
05.06.2008 27
Steps in Simulation Study
Problem Formulation Setting objectives & Plan Model Conceptualization Data Collection Model Translation Verify model Validate model Experimental Design
05.06.2008 28
Fundamentally an iterative process
Steps in Simulation Study
From previous slide Production run & Analysis More runs?
Documentation & Reporting
Implementation
05.06.2008 29
Problem Formulation
Initial step Identify controllable and uncontrollable inputs Identify constraints on the decision variables Define measure of system performance and an objective function Develop a preliminary model structure to interrelate the inputs and the measure of performance May be the problem needs reformulation as the study progresses
05.06.2008 30
Setting Objectives & Plan
What do you (or the customer) hope to accomplish with the model
May be an end in itself Predict the weather Train personnel to develop certain skills (e.g., driving) More often a means to an end Optimize a manufacturing process or develop the most cost effective means to reduce traffic congestion in some part of a city
Often requires developing a business case to justify the cost
Improved efficiency will save the company $$$ Example: electronics Even so, may be hard to justify in lean times
Goals may not be known when you start the project!
One often learns things along the way
05.06.2008 31
Develop Conceptual Model
An abstract (i.e., not directly executable) representation
of the system What should be included in model? What can be left out? What abstractions should be used
Level of detail Often a variation on standard abstractions Example: transportation
Fluid flow? Queuing network? Cellular automation?
What metrics will be produced by the model? Appropriate choice depends on the purpose of the model
05.06.2008 32
Data Collection
Regardless of the method used to collect the data, the
decision of how much to collect is a trade-off between cost and accuracy Constant inter play between construction of the model and the collection of needed input Changes with the degree of complexity of the model Data should be collected for the validation as well
05.06.2008
33
Model translation
Model requires great deal of information and
computation Needs to be translated into computer recognizable format using either special purpose or general purpose languages Focus of this course will be using Excel for model building Arena characteristics will be introduced
05.06.2008
34
Verification & Validation
Verification focuses on the internal consistency of a
model Validation is concerned with the correspondence between the model and the reality Validation is applied to those processes which seek to determine whether or not a simulation is correct with respect to the "real" system Validation is concerned with the question "Are we building the right system?
05.06.2008
35
Verification & Validation
Verification seeks to answer the question "Are we
building the system right?" Verification checks that the implementation of the simulation model (program) corresponds to the model Validation checks that the model corresponds to reality Calibration checks that the data generated by the simulation matches real (observed) data.
05.06.2008
36
Experimental Design
Alternatives to be simulated must be determined Good experimental design Randomization Replication Local control For each system decisions needed Length of the initialization period Length of the simulation run Number of replication
05.06.2008
37
Production runs and analysis
To measure performance of the simulation system so
designed Also to determine if more runs needed till results are consistent
05.06.2008
38
Documentation & Reporting
Two types Program
Needed if it is to be used again May need to be applied for different system by different people For modification Provides important written history of simulation project Should be frequent as the project progresses
Progress
05.06.2008
39
Implementation
Success depends how well previous steps were
followed .
05.06.2008
40
M S Prasad : An avid researcher in the field of
Signal processing & Image processing, applicable to defence system. He has been designer of on Board computer Navigation system, Multiple Target tracking for missiles he started the digital GIS system in the country and has been responsible for having a total digital military Ops room . In his span of career , he has been Network security auditor for UNO classified organisations. He is the inventor most secure strategic system for India alongwith Positive Activation& Safing System.(PASS) He holds 4 patents and has been awarded twice the Best scientist by defence Research organisation. Lately he is the chief evangelist of Cloud Security in ASIA. He is recipient of numerous national & International awards. He has published 28 papers in referred journals and 3 in Monographs
05.06.2008 41
Thanks
05.06.2008
42
