Fundamental

Simulation
Concepts

Chapter 2

Last revision August 12, 2006

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 1 of 57
What We’ll Do ...
• Underlying ideas, methods, and issues in
simulation
• Software-independent (setting up for Arena)
• Example of a simple processing system
 Decompose the problem
 Terminology
 Simulation by hand
 Some basic statistical issues
• Spreadsheet simulation
 Simple static, dynamic models
• Overview of a simulation study
Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 2 of 57
The System:
A Simple Processing System
Machine
(Server)
Arriving Departing
7 6 5 4
Blank Parts Finished Parts
Queue (FIFO) Part in Service
• General intent:
 Estimate expected production
 Waiting time in queue, queue length, proportion of time
machine is busy
• Time units
 Can use different units in different places … must declare
 Be careful to check the units when specifying inputs
 Declare base time units for internal calculations, outputs
 Be reasonable (interpretation, roundoff error)
Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 3 of 57
Model Specifics
• Initially (time 0) empty and idle
• Base time units: minutes
• Input data (assume given for now …), in minutes:
Part Number Arrival Time Interarrival Time Service Time
1 0.00 1.73 2.90
2 1.73 1.35 1.76
3 3.08 0.71 3.39
4 3.79 0.62 4.52
5 4.41 14.28 4.46
6 18.69 0.70 4.36
7 19.39 15.52 2.07
8 34.91 3.15 3.36
9 38.06 1.76 2.37
10 39.82 1.00 5.38
11 40.82 . .
. . . .
. . . .
• Stop when 20 minutes of (simulated) time have
passed
Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 4 of 57
Goals of the Study:
Output Performance Measures
• Total production of parts over the run (P)
• Average waiting time of parts in queue:
N N = no. of parts completing queue wait
 WQi WQi = waiting time in queue of ith part
i 1
N Know: WQ1 = 0 (why?)
N > 1 (why?)

• Maximum waiting time of parts in queue:

max WQi
i 1,...,N

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 5 of 57
Goals of the Study:
Output Performance Measures (cont’d.)
• Time-average number of parts in queue:
20
0 Q(t ) dt Q(t) = number of parts in queue
20 at time t
max Q (t )
• Maximum number of parts in queue: 0t 20
• Average and maximum total time in system of
parts (a.k.a. cycle time):
P
 TSi
i 1 , max TSi TSi = time in system of part i
P i 1,...,P

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 6 of 57
Goals of the Study:
Output Performance Measures (cont’d.)
• Utilization of the machine (proportion of time
busy)
20
0 B(t ) dt 1 if the machine is busy at time t
, B(t )  
20 0 if the machine is idle at time t

• Many others possible (information overload?)

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 7 of 57
Analysis Options
• Educated guessing
 Average interarrival time = 4.08 minutes
 Average service time = 3.46 minutes
 So (on average) parts are being processed faster than they
arrive
– System has a chance of operating in a stable way in the long run,
i.e., might not “explode”
– If all interarrivals and service times were exactly at their mean,
there would never be a queue
– But the data clearly exhibit variability, so a queue could form
 If we’d had average interarrival < average service time, and
this persisted, then queue would explode
 Truth — between these extremes
 Guessing has its limits …
Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 8 of 57
Analysis Options (cont’d.)
• Queueing theory
 Requires additional assumptions about the model
 Popular, simple model: M/M/1 queue
– Interarrival times ~ exponential
– Service times ~ exponential, indep. of interarrivals
– Must have E(service) < E(interarrival)
– Steady-state (long-run, forever)
– Exact analytic results; e.g., average waiting time in queue is
S2  A  E(interarrival time)
,
 A  S S  E(service time)
 Problems: validity, estimating means, time frame
 Often useful as first-cut approximation

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 9 of 57
Mechanistic Simulation
• Individual operations (arrivals, service times) will
occur exactly as in reality
• Movements, changes occur at the right “time,” in
the right order
• Different pieces interact
• Install “observers” to get output performance
measures
• Concrete, “brute-force” analysis approach
• Nothing mysterious or subtle
 But a lot of details, bookkeeping
 Simulation software keeps track of things for you
Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 10 of 57
Pieces of a Simulation Model
• Entities
 “Players” that move around, change status, affect and are
affected by other entities
 Dynamic objects — get created, move around, leave
(maybe)
 Usually represent “real” things
– Our model: entities are the parts
 Can have “fake” entities for modeling “tricks”
– Breakdown demon, break angel
Though Arena has built-in ways to model these examples directly
 Usually have multiple realizations floating around
 Can have different types of entities concurrently
 Usually, identifying the types of entities is the first thing to do
in building a model
Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 11 of 57
Pieces of a Simulation Model (cont’d.)
• Attributes
 Characteristic of all entities: describe, differentiate
 All entities have same attribute “slots” but different values
for different entities, for example:
– Time of arrival
– Due date
– Priority
– Color
 Attribute value tied to a specific entity
 Like “local” (to entities) variables
 Some automatic in Arena, some you define

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 12 of 57
Pieces of a Simulation Model (cont’d.)
• (Global) Variables
 Reflects a characteristic of the whole model, not of specific
entities
 Used for many different kinds of things
– Travel time between all station pairs
– Number of parts in system
– Simulation clock (built-in Arena variable)
 Name, value of which there’s only one copy for the whole
model
 Not tied to entities
 Entities can access, change variables
 Writing on the wall (rewriteable)
 Some built-in by Arena, you can define others

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 13 of 57
Pieces of a Simulation Model (cont’d.)
• Resources
 What entities compete for
– People
– Equipment
– Space
 Entity seizes a resource, uses it, releases it
 Think of a resource being assigned to an entity, rather than
an entity “belonging to” a resource
 “A” resource can have several units of capacity
– Seats at a table in a restaurant
– Identical ticketing agents at an airline counter
 Number of units of resource can be changed during the
simulation

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 14 of 57
Pieces of a Simulation Model (cont’d.)
• Queues
 Place for entities to wait when they can’t move on (maybe
since the resource they want to seize is not available)
 Have names, often tied to a corresponding resource
 Can have a finite capacity to model limited space — have
to model what to do if an entity shows up to a queue that’s
already full
 Usually watch the length of a queue, waiting time in it

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 15 of 57
Pieces of a Simulation Model (cont’d.)
• Statistical accumulators
 Variables that “watch” what’s happening
 Depend on output performance measures desired
 “Passive” in model — don’t participate, just watch
 Many are automatic in Arena, but some you may have to
set up and maintain during the simulation
 At end of simulation, used to compute final output
performance measures

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 16 of 57
Pieces of a Simulation Model (cont’d.)
• Statistical accumulators for the simple
processing system
 Number of parts produced so far
 Total of the waiting times spent in queue so far
 No. of parts that have gone through the queue
 Max time in queue we’ve seen so far
 Total of times spent in system
 Max time in system we’ve seen so far
 Area so far under queue-length curve Q(t)
 Max of Q(t) so far
 Area so far under server-busy curve B(t)

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 17 of 57
Simulation Dynamics:
The Event-Scheduling “World View”
• 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.)
Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 18 of 57
Events for the
Simple Processing System
• Arrival of a new part to the system
 Update time-persistent statistical accumulators (from last
event to now)
– Area under Q(t)
– Max of Q(t)
– Area under B(t)
 “Mark” arriving part with current time (use later)
 If machine is idle:
– Start processing (schedule departure), Make machine busy, Tally
waiting time in queue (0)
 Else (machine is busy):
– Put part at end of queue, increase queue-length variable
 Schedule the next arrival event
Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 19 of 57
Events for the
Simple Processing System (cont’d.)
• Departure (when a service is completed)
 Increment number-produced stat accumulator
 Compute & tally time in system (now - time of arrival)
 Update time-persistent statistics (as in arrival event)
 If queue is non-empty:
– Take first part out of queue, compute & tally its waiting time in
queue, begin service (schedule departure event)
 Else (queue is empty):
– Make the machine idle (Note: there will be no departure event
scheduled on the future events calendar, which is as desired)

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 20 of 57
Events for the
Simple Processing System (cont’d.)
• The End
 Update time-persistent statistics (to end of the simulation)
 Compute final output performance measures using current
(= final) values of statistical accumulators
• After each event, the event calendar’s top record
is removed to see what time it is, what to do
• Also must initialize everything

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 21 of 57
Some Additional Specifics for the
Simple Processing System
• Simulation clock variable (internal in Arena)
• Event calendar: list of event records:
 [Entity No., Event Time, Event Type]
 Keep ranked in increasing order on Event Time
 Next event always in top record
 Initially, schedule first Arrival, The End (Dep.?)
• State variables: describe current status
 Server status B(t) = 1 for busy, 0 for idle
 Number of customers in queue Q(t)
 Times of arrival of each customer now in queue (a list of
random length)

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 22 of 57
Simulation by Hand
• Manually track state variables, statistical
accumulators
• Use “given” interarrival, service times
• Keep track of event calendar
• “Lurch” clock from one event to the next
• Will omit times in system, “max” computations
here (see text for complete details)

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 23 of 57
Simulation by Hand:
Setup
System Clock B(t) Q(t) Arrival times of Event calendar
custs. in queue

Number of Total of Area under Area under

completed waiting waiting times in queue Q(t) B(t)
times in queue

4
3
Q(t) graph 2
1
0
0 5 10 15 20
2
B(t) graph 1
0
0 5 10 15 20

Time (Minutes)
Interarrival times 1.73, 1.35, 0.71, 0.62, 14.28, 0.70, 15.52, 3.15, 1.76, 1.00, ...
Service times 2.90, 1.76, 3.39, 4.52, 4.46, 4.36, 2.07, 3.36, 2.37, 5.38, ...

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 24 of 57
Simulation by Hand:
t = 0.00, Initialize
System Clock B(t) Q(t) Arrival times of Event calendar
custs. in queue [1, 0.00, Arr]
0.00 0 0 <empty> [–, 20.00, End]

Number of Total of Area under Area under

completed waiting waiting times in queue Q(t) B(t)
times in queue
0 0.00 0.00 0.00

4
3
Q(t) graph 2
1
0
0 5 10 15 20
2
B(t) graph 1
0
0 5 10 15 20

Time (Minutes)
Interarrival times 1.73, 1.35, 0.71, 0.62, 14.28, 0.70, 15.52, 3.15, 1.76, 1.00, ...
Service times 2.90, 1.76, 3.39, 4.52, 4.46, 4.36, 2.07, 3.36, 2.37, 5.38, ...

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 25 of 57
Simulation by Hand:
t = 0.00, Arrival of Part 1
System Clock B(t) Q(t) Arrival times of Event calendar
custs. in queue [2, 1.73, Arr]
1 0.00 1 0 <empty> [1, 2.90, Dep]
[–, 20.00, End]
Number of Total of Area under Area under
completed waiting waiting times in queue Q(t) B(t)
times in queue
1 0.00 0.00 0.00

4
3
Q(t) graph 2
1
0
0 5 10 15 20
2
B(t) graph 1
0
0 5 10 15 20

Time (Minutes)
Interarrival times 1.73, 1.35, 0.71, 0.62, 14.28, 0.70, 15.52, 3.15, 1.76, 1.00, ...
Service times 2.90, 1.76, 3.39, 4.52, 4.46, 4.36, 2.07, 3.36, 2.37, 5.38, ...

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 26 of 57
Simulation by Hand:
t = 1.73, Arrival of Part 2
System Clock B(t) Q(t) Arrival times of Event calendar
custs. in queue [1, 2.90, Dep]
2 1 1.73 1 1 (1.73) [3, 3.08, Arr]
[–, 20.00, End]
Number of Total of Area under Area under
completed waiting waiting times in queue Q(t) B(t)
times in queue
1 0.00 0.00 1.73

4
3
Q(t) graph 2
1
0
0 5 10 15 20
2
B(t) graph 1
0
0 5 10 15 20

Time (Minutes)
Interarrival times 1.73, 1.35, 0.71, 0.62, 14.28, 0.70, 15.52, 3.15, 1.76, 1.00, ...
Service times 2.90, 1.76, 3.39, 4.52, 4.46, 4.36, 2.07, 3.36, 2.37, 5.38, ...

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 27 of 57
Simulation by Hand:
t = 2.90, Departure of Part 1
System Clock B(t) Q(t) Arrival times of Event calendar
custs. in queue [3, 3.08, Arr]
2 2.90 1 0 <empty> [2, 4.66, Dep]
[–, 20.00, End]
Number of Total of Area under Area under
completed waiting waiting times in queue Q(t) B(t)
times in queue
2 1.17 1.17 2.90

4
3
Q(t) graph 2
1
0
0 5 10 15 20
2
B(t) graph 1
0
0 5 10 15 20

Time (Minutes)
Interarrival times 1.73, 1.35, 0.71, 0.62, 14.28, 0.70, 15.52, 3.15, 1.76, 1.00, ...
Service times 2.90, 1.76, 3.39, 4.52, 4.46, 4.36, 2.07, 3.36, 2.37, 5.38, ...

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 28 of 57
Simulation by Hand:
t = 3.08, Arrival of Part 3
System Clock B(t) Q(t) Arrival times of Event calendar
custs. in queue [4, 3.79, Arr]
3 2 3.08 1 1 (3.08) [2, 4.66, Dep]
[–, 20.00, End]
Number of Total of Area under Area under
completed waiting waiting times in queue Q(t) B(t)
times in queue
2 1.17 1.17 3.08

4
3
Q(t) graph 2
1
0
0 5 10 15 20
2
B(t) graph 1
0
0 5 10 15 20

Time (Minutes)
Interarrival times 1.73, 1.35, 0.71, 0.62, 14.28, 0.70, 15.52, 3.15, 1.76, 1.00, ...
Service times 2.90, 1.76, 3.39, 4.52, 4.46, 4.36, 2.07, 3.36, 2.37, 5.38, ...

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 29 of 57
Simulation by Hand:
t = 3.79, Arrival of Part 4
System Clock B(t) Q(t) Arrival times of Event calendar
custs. in queue [5, 4.41, Arr]
4 3 2 3.79 1 2 (3.79, 3.08) [2, 4.66, Dep]
[–, 20.00, End]
Number of Total of Area under Area under
completed waiting waiting times in queue Q(t) B(t)
times in queue
2 1.17 1.88 3.79

4
3
Q(t) graph 2
1
0
0 5 10 15 20
2
B(t) graph 1
0
0 5 10 15 20

Time (Minutes)
Interarrival times 1.73, 1.35, 0.71, 0.62, 14.28, 0.70, 15.52, 3.15, 1.76, 1.00, ...
Service times 2.90, 1.76, 3.39, 4.52, 4.46, 4.36, 2.07, 3.36, 2.37, 5.38, ...

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 30 of 57
Simulation by Hand:
t = 4.41, Arrival of Part 5
System Clock B(t) Q(t) Arrival times of Event calendar
custs. in queue [2, 4.66, Dep]
5 4 3 2 4.41 1 3 (4.41, 3.79, 3.08) [6, 18.69, Arr]
[–, 20.00, End]
Number of Total of Area under Area under
completed waiting waiting times in queue Q(t) B(t)
times in queue
2 1.17 3.12 4.41

4
3
Q(t) graph 2
1
0
0 5 10 15 20
2
B(t) graph 1
0
0 5 10 15 20

Time (Minutes)
Interarrival times 1.73, 1.35, 0.71, 0.62, 14.28, 0.70, 15.52, 3.15, 1.76, 1.00, ...
Service times 2.90, 1.76, 3.39, 4.52, 4.46, 4.36, 2.07, 3.36, 2.37, 5.38, ...

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 31 of 57
Simulation by Hand:
t = 4.66, Departure of Part 2
System Clock B(t) Q(t) Arrival times of Event calendar
custs. in queue [3, 8.05, Dep]
5 4 3 4.66 1 2 (4.41, 3.79) [6, 18.69, Arr]
[–, 20.00, End]
Number of Total of Area under Area under
completed waiting waiting times in queue Q(t) B(t)
times in queue
3 2.75 3.87 4.66

4
3
Q(t) graph 2
1
0
0 5 10 15 20
2
B(t) graph 1
0
0 5 10 15 20

Time (Minutes)
Interarrival times 1.73, 1.35, 0.71, 0.62, 14.28, 0.70, 15.52, 3.15, 1.76, 1.00, ...
Service times 2.90, 1.76, 3.39, 4.52, 4.46, 4.36, 2.07, 3.36, 2.37, 5.38, ...

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 32 of 57
Simulation by Hand:
t = 8.05, Departure of Part 3
System Clock B(t) Q(t) Arrival times of Event calendar
custs. in queue [4, 12.57, Dep]
5 4 8.05 1 1 (4.41) [6, 18.69, Arr]
[–, 20.00, End]
Number of Total of Area under Area under
completed waiting waiting times in queue Q(t) B(t)
times in queue
4 7.01 10.65 8.05

4
3
Q(t) graph 2
1
0
0 5 10 15 20
2
B(t) graph 1
0
0 5 10 15 20

Time (Minutes)
Interarrival times 1.73, 1.35, 0.71, 0.62, 14.28, 0.70, 15.52, 3.15, 1.76, 1.00, ...
Service times 2.90, 1.76, 3.39, 4.52, 4.46, 4.36, 2.07, 3.36, 2.37, 5.38, ...

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 33 of 57
Simulation by Hand:
t = 12.57, Departure of Part 4
System Clock B(t) Q(t) Arrival times of Event calendar
custs. in queue [5, 17.03, Dep]
5 12.57 1 0 () [6, 18.69, Arr]
[–, 20.00, End]
Number of Total of Area under Area under
completed waiting waiting times in queue Q(t) B(t)
times in queue
5 15.17 15.17 12.57

4
3
Q(t) graph 2
1
0
0 5 10 15 20
2
B(t) graph 1
0
0 5 10 15 20

Time (Minutes)
Interarrival times 1.73, 1.35, 0.71, 0.62, 14.28, 0.70, 15.52, 3.15, 1.76, 1.00, ...
Service times 2.90, 1.76, 3.39, 4.52, 4.46, 4.36, 2.07, 3.36, 2.37, 5.38, ...

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 34 of 57
Simulation by Hand:
t = 17.03, Departure of Part 5
System Clock B(t) Q(t) Arrival times of Event calendar
custs. in queue [6, 18.69, Arr]
17.03 0 0 () [–, 20.00, End]

Number of Total of Area under Area under

completed waiting waiting times in queue Q(t) B(t)
times in queue
5 15.17 15.17 17.03

4
3
Q(t) graph 2
1
0
0 5 10 15 20
2
B(t) graph 1
0
0 5 10 15 20

Time (Minutes)
Interarrival times 1.73, 1.35, 0.71, 0.62, 14.28, 0.70, 15.52, 3.15, 1.76, 1.00, ...
Service times 2.90, 1.76, 3.39, 4.52, 4.46, 4.36, 2.07, 3.36, 2.37, 5.38, ...

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 35 of 57
Simulation by Hand:
t = 18.69, Arrival of Part 6
System Clock B(t) Q(t) Arrival times of Event calendar
custs. in queue [7, 19.39, Arr]
6 18.69 1 0 () [–, 20.00, End]
[6, 23.05, Dep]
Number of Total of Area under Area under
completed waiting waiting times in queue Q(t) B(t)
times in queue
6 15.17 15.17 17.03

4
3
Q(t) graph 2
1
0
0 5 10 15 20
2
B(t) graph 1
0
0 5 10 15 20

Time (Minutes)
Interarrival times 1.73, 1.35, 0.71, 0.62, 14.28, 0.70, 15.52, 3.15, 1.76, 1.00, ...
Service times 2.90, 1.76, 3.39, 4.52, 4.46, 4.36, 2.07, 3.36, 2.37, 5.38, ...

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 36 of 57
Simulation by Hand:
t = 19.39, Arrival of Part 7
System Clock B(t) Q(t) Arrival times of Event calendar
custs. in queue [–, 20.00, End]
7 6 19.39 1 1 (19.39) [6, 23.05, Dep]
[8, 34.91, Arr]
Number of Total of Area under Area under
completed waiting waiting times in queue Q(t) B(t)
times in queue
6 15.17 15.17 17.73

4
3
Q(t) graph 2
1
0
0 5 10 15 20
2
B(t) graph 1
0
0 5 10 15 20

Time (Minutes)
Interarrival times 1.73, 1.35, 0.71, 0.62, 14.28, 0.70, 15.52, 3.15, 1.76, 1.00, ...
Service times 2.90, 1.76, 3.39, 4.52, 4.46, 4.36, 2.07, 3.36, 2.37, 5.38, ...

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 37 of 57
Simulation by Hand:
t = 20.00, The End
System Clock B(t) Q(t) Arrival times of Event calendar
custs. in queue [6, 23.05, Dep]
7 6 20.00 1 1 (19.39) [8, 34.91, Arr]

Number of Total of Area under Area under

completed waiting waiting times in queue Q(t) B(t)
times in queue
6 15.17 15.78 18.34

4
3
Q(t) graph 2
1
0
0 5 10 15 20
2
B(t) graph 1
0
0 5 10 15 20

Time (Minutes)
Interarrival times 1.73, 1.35, 0.71, 0.62, 14.28, 0.70, 15.52, 3.15, 1.76, 1.00, ...
Service times 2.90, 1.76, 3.39, 4.52, 4.46, 4.36, 2.07, 3.36, 2.37, 5.38, ...

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 38 of 57
Simulation by Hand:
Finishing Up
• Average waiting time in queue:
Total of times in queue 15.17
  2.53 minutes per part
No. of times in queue 6

• Time-average number in queue:

Area under Q(t ) curve 15.78
  0.79 part
Final clock value 20

• Utilization of drill press:

Area under B(t ) curve 18.34
  0.92 (dimensionless)
Final clock value 20

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 39 of 57
Complete Record of the
Hand Simulation

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 40 of 57
Event-Scheduling Logic via
Programming
• Clearly well suited to standard programming
language
• Often use “utility” libraries for:
 List processing
 Random-number generation
 Random-variate generation
 Statistics collection
 Event-list and clock management
 Summary and output
• Main program ties it together, executes events in
order
Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 41 of 57
Simulation Dynamics:
The Process-Interaction World View
• Identify characteristic entities in the system
• Multiple copies of entities co-exist, interact,
compete
• “Code” is non-procedural
• 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
 Arena translates your model description into a program in
the SIMAN simulation language for execution
Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 42 of 57
Randomness in Simulation
• The above was just one “replication” — a sample of
size one (not worth much)
• Made a total of five replications (IID):

Note
substantial
variability
across
replications

• Confidence intervals
X  tn 1,1for
/ 2s /expected
n values:
 In general, (normality assumption?)
3.80  (2.776 )(1.64 / 5 )
 For expected total production, 3.80  2.04
Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 43 of 57
Comparing Alternatives
• Usually, simulation is used for more than just a
single model “configuration”
• Often want to compare alternatives, select or
search for the best (via some criterion)
• Simple processing system: What would happen
if the arrival rate were to double?
 Cut interarrival times in half
 Rerun the model for double-time arrivals
 Make five replications

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 44 of 57
Results: Original vs. Double-Time
Arrivals
• Original – circles
• Double-time – triangles
• Replication 1 – filled in
• Replications 2-5 – hollow
• Note variability
• Danger of making decisions
based on one (first)
replication
• Hard to see if there are
really differences
• Need: Statistical analysis of
simulation output data
Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 45 of 57
Simulating with Spreadsheets:
Introduction
• Popular, ubiquitous tool
• Can use for simple simulation models
 Typically, static models
– Risk analysis, financial/investment scenarios
 Only the simplest of dynamic models
• Two examples
 News vendor problem
 Waiting times in single-server queue
– Special recursion valid only in this case

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 46 of 57
Simulating with Spreadsheets:
Newsvendor Problem – Setup
• Newsvendor sells newspapers on the street
 Buys for c = $0.55 each, sells for r = $1.00 each
• Each morning, buys q copies
 q is a fixed number, same every day
• Demand during a day: D = max (X, 0)
 X ~ normal ( = 135.7,  = 27.1), from historical data
 X rounds X to nearest integer
• If D  q, satisfy all demand, and q – D  0 left over,
sell for scrap at s = $0.03 each
• If D > q, sells out (sells all q copies), no scrap
 But missed out on D – q > 0 sales
• What should q be?
Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 47 of 57
Simulating with Spreadsheets:
Newsvendor Problem – Formulation
• Choose q to maximize expected profit per day
 q too small – sell out, miss $0.45 profit per paper
 q too big – have left over, scrap at a loss of $0.52 per paper
• Classic operations-research problem
 Many versions, variants, extensions, applications
 Much research on exact solution in certain cases
 But easy to simulate, even in a spreadsheet
• Profit in a day, as a function of q:
W(q) = r min (D, q) + s max (q – D, 0) – cq
Sales revenue Scrap revenue Cost
 W(q) is a random variable – profit varies from day to day
• Maximize E(W(q)) over nonnegative integers q
Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 48 of 57
Simulating with Spreadsheets:
Newsvendor Problem – Simulation
• Set trial value of q, generate demand D, compute
profit for that day
 Then repeat this for many days independently, average to
estimate E(W(q))
– Also get confidence interval, estimate of P(loss), histogram of W(q)
 Try for a range of values of q
• Need to generate demand D = max (X, 0)
 So need to generate X ~ normal ( = 135.7,  = 27.1)
 (Much) ahead – Sec. 12.2, generating random variates
1
 In this case, generate X = ,(U)
U is a random number distributed uniformly on [0, 1] (Sec. 12.1)
, is cumulative distribution function of normal (, ) distribtuion
Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 49 of 57
Simulating with Spreadsheets:
Newsvendor Problem – Excel
• File Newsvendor.xls
• Input parameters in cells B4 – B8 (blue)
• Trial values for q in row 2 (pink)
• Day number (1, 2, ..., 30) in column D
• Demands in column E for each day:
= MAX(ROUND(NORMINV(RAND(), $B$7, $B$8), 0), 0)
Rounding 1 U(0, 1)  
function random number

RAND() is “volatile” X ~ normal (, )

so regenerates on Round to
any edit, or F9 key $ pins down following nearest
MAX 2nd
column or row when integer
argument
copying
Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 50 of 57
Simulating with Spreadsheets:
Newsvendor Problem – Excel (cont’d.)
• For each q:
 “Sold” column: number of papers sold that day
 “Scrap” column: number of papers scrapped that day
 “Profit” column: profit (+, –, 0) that day
 Placement of “$” in formulas to facilitate copying
• At bottom of “Profit” columns (green):
 Average profit over 30 days
 Half-width of 95% confidence interval on E(W(q))
– Value 2.045 is upper 0.975 critical point of t distribution with 29 d.f.
– Plot confidence intervals as “I-beams” on left edge
 Estimate of P(W(q) < 0)
– Uses COUNTIF function
• Histograms of W(q) at bottom
 Vertical red line at 0, separates profits, losses
Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 51 of 57
Simulating with Spreadsheets:
Newsvendor Problem – Results
• Fine point – used same daily demands (column E)
for each day, across all trial values of q
 Would have been valid to generate them independently
 Why is it better to use the same demands for all q?
• Results
 Best q is about 140, maybe a little less
 Randomness in all the results (tap F9 key)
– All demands, profits, graphics change
– Confidence-interval, histogram plots change
– Reminder that these are random outputs, random plots
 Higher q  more variability in profit
– Histograms at bottom are wider for larger q
– Higher chance of both large profits, but higher chance of loss, too
– Risk/return tradeoff can be quantified – risk taker vs. risk-averse
Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 52 of 57
Simulating with Spreadsheets:
Single-Server Queue – Setup
• Like hand simulation, but
 Interarrival times ~ exponential with mean 1/  = 1.6 min.
 Service times ~ uniform on [a, b] = [0.27, 2.29] min.
 Stop when 50th waiting time in queue is observed
– i.e., when 50th customer begins service, not exits system

• Watch waiting times in queue WQ1, WQ2, ..., WQ50

 Important – not watching anything else, unlike before
• Si = service time of customer i,
Ai = interarrival time between custs. i – 1 and i
• Lindley’s recursion (1952): Initialize WQ1 = 0,
WQi = max (WQi – 1 + Si – 1 – Ai, 0), i = 2, 3, ...
Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 53 of 57
Simulating with Spreadsheets:
Single-Server Queue – Simulation
• Need to generate random variates: let U ~ U[0, 1]
 Exponential (mean 1/): Ai = –(1/) ln(1 – U)
 Uniform on [a, b]: Si = a + (b – a) U
• File MU1.xls
• Input parameters in cells B4 – B6 (blue)
 Some theoretical outputs in cells B8 – B10
• Customer number (i = 1, 2, ..., 50) in column D
• Five IID replications (three columns for each)
 IA = interarrival times, S = service times
 WQ = waiting times in queue (plot, thin curves)
– First one initialized to 0, remainder use Lindley’s recursion
Curves rise from 0, variation increases toward right
– Creates positive autocorrelation down the WQ columns
Curves have less abrupt jumps than if WQi’s were independent

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 54 of 57
Simulating with Spreadsheets:
Single-Server Queue – Results
• Column averages (green)
 Average interarrival, service times close to expectations
 Average WQi within each replication
– Not too far from steady-state expectation
– Considerable variation
– Many are below it (why?)
• Cross-replication (by customer) averages (green)
 Column T, thick line in plot to dampen noise
• Why no sample variance, histograms of WQi’s?
 Could have computed both, as in newsvendor
 Nonstationarity – what is a “typical” WQi here?
 Autocorrelation – biases variance estimate, may bias
histogram if run is not “long enough”
Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 55 of 57
Simulating with Spreadsheets:
Recap
• Popular for static models
 Add-ins – @RISK, Crystal Ball
• Inadequate tool for dynamic simulations if there’s
any complexity
 Extremely easy to simulate the single-server queue in
Arena – Chapter 3 main example
 Can build very complex dynamic models with Arena – most
of the rest of the book

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 56 of 57
Overview of a Simulation Study
• Understand the system
• Be clear about the goals
• Formulate the model representation
• Translate into modeling software
• Verify “program”
• Validate model
• Design experiments
• Make runs
• Analyze, get insight, document results

Simulation with Arena, 4th ed. Chapter 2 – Fundamental Simulation Concepts Slide 57 of 57