airport landside modeling (computer applications and...
TRANSCRIPT
Virginia Polytechnic Institute and State University 1 of 77
Airport Landside Modeling(Computer Applications and Modeling)
Dr. Antonio A. TraniProfessor of Civil and Environmental Engineering Virginia Polytechnic Institute and State University
CEE 4674 - Airport Planning and Design
Virginia Polytechnic Institute and State University 2 of 77
Material Presented in this Section
• Brief description of airport terminal simulationlanguages
• Advantages and weaknesses
• Basic constructs
• Example of VPI Airport Terminal SimulationLibrary (EXTEND)
• Example APM simulation model (APMSIM)developed at Virginia Tech
• Future directions and impacts
Virginia Polytechnic Institute and State University 3 of 77
Purpose of the Discussion
• Until now you understand the basics of probability,continuous, and discrete event simulation and modelingusing generic mathematical packages (such as Matlab) orspreadsheets
• Dedicated simulation languages keep track of many ofthe book keeping activities required in the simulationmodel
• you can concentrate in the model with minimalimplementation burden
• productivity and model development are enhanced
• Simulation tools provide rich GUI constructs to enhanceunderstanding from a decision maker viewpoint
Virginia Polytechnic Institute and State University 4 of 77
Simulation Languages
Discrete Event Simulation Languages
SIMULA, STELLA, SIMSCRIPT II.5, MODSIM, SLAM III, GPSS-H/GPSS-PC, Arena/SIMAN, EXTEND
Continuous Simulation Languages
ACSL, Simulink, STELLA, SIMSCRIPT II.5, MODSIM, SLAM III, GPSS-H/GPSS-PC, Arena/SIMAN, EXTEND
• Many of these languages provide a hybrid modelingcapability (i.e., discrete and continuous paradigms in thesame model)
Virginia Polytechnic Institute and State University 5 of 77
A Typical Simulation Language
• While all languages provide very similar constructs to build models, some have more complex Graphic User Interfaces (GUI) than others
• The basic building blocks of a simulation model are maintained.
• Building blocks of discrete simulation languages:
• Entities, Attributes, Resources, Queues, Accumulators, and Events
• Building blocks of continuous simulation languages:
• Integrator (reservoir, tank), generator, delay structures, function blocks, rate variable blocks, etc.
Virginia Polytechnic Institute and State University 6 of 77
A Typical Simulation Language (cont.)
• Modern (3rd generation simulation languages) provide rich GUI interfaces to construct and prototype models (EXTEND, Stella, and Arena are good examples of these)
• Object-oriented data abstraction is the norm
• All simulation languages have connectivity capabilities with other packages (i.e., statistical, spreadsheets, word processors, programming)
• ODBC/Corba support for database connectivity
• OLE support for Windows programming compatibility
• Statistical package support for regression and input/output analysis
Virginia Polytechnic Institute and State University 7 of 77
Example of a Simulation Language Use and Connectivity
SimulationLanguageEngine
Aviation Data
Statistical
Packageor Numerical
Database
Package
SimulationLanguageGraphics
or GISDecisionSupportModel
Corba
OLECorba
Corba
Virginia Polytechnic Institute and State University 8 of 77
Sample 2nd Generation Sim. Language (Simscript)
• Developed and marketed by CACI (US)
• Provides English-type constructs
• C-type model construction
• Preamble file (contains global variables, resourcedefinitions, event types, tally definitions, etc.)
• Main file (first routine to be executed: calls othersin the program)
• C-language portability (platform independence)
• Heavily used in military and aviation modeling in the1970’s and 1980’s in US
• Stable compiler (few modifications in the past five years)
Virginia Polytechnic Institute and State University 9 of 77
Typical Organization of a Simscript II Model
Preamble
variable definitionsprocess definitionsresource definitions
Main
call process 1local var. definitions
call process m
Process 1
code to accomplish
local var. definitions
call process ma task
Process m
code to accomplish
local var. definitions
call process ma task
Virginia Polytechnic Institute and State University 10 of 77
Sample 2nd Generation Sim. Language (Simscript)
• Graphic capabilities were added in the 1980’s (Simgraphics)
• One of the first simulation languages used in distributed simulation mode (i.e., running processes in different computers)
• SIMMOD - the airspace and airfield simulation model developed by the FAA was developed in Simscript II.5 in the late 1970’s
• Today, MODSIM is replacing Simscript II.5 and adding full Object-Oriented Programming paradigm capabilities (RAMS - another airspace simulation model by Eurocontrol was developed in using MODSIM)
Virginia Polytechnic Institute and State University 11 of 77
Sample 3rd Generation Sim. Language (EXTEND)
• Developed by Bod Diamond (1987) and distributed by Imagine That, Inc.
• Provides a rich GUI to develop models using blocks
• Access to C-like block code (Modl language)
• Blocks are classified in terms of libraries
• Drag-and-connect approach
• Hierarchy blocks (multiple blocks working towards a common goal)
• Some portability (WIN and Mac versions available)
• Used im manufacturing, transportation and control engineering
Virginia Polytechnic Institute and State University 12 of 77
Sample 3rd Generation Sim. Language (EXTEND)
Extend Libraries
Generic Library: has blocks that perform basic functions as math, decision handling and input/output
Discrete-Event Library: contains blocks for creating discrete-event simulation models (including activity delays, resources, processes, etc.)
Plotter Library: contains blocks for plotting the results in any type of simulation model
Animation Library: holds blocks that are used for animating hierarchical blocks
Virginia Polytechnic Institute and State University 13 of 77
Taxonomy of a Block
Blocks are complex structures in Extend that allow quick modeling of complex processes. Shown below is a typical block showing its components.
F
L W
1
Uni-queue Line
Input Connector
Output Connector
Information Connector(Output Connector)
Information Connector(Output Connector)
Virginia Polytechnic Institute and State University 14 of 77
Taxonomy of a Block (cont.)
Each block in Extend is extensible and modifiable to the source code level (Modl - the language used in Extend - blocks is an extension of the C language).
procedure getArrays()
{
if (rCount != sysGlobalint2)
{
if(useString)
getPassedArray(sysGlobal5, itemArrayA);
if (costing)
getPassedArray(sysGlobal9, itemArrayC);
rCount = sysGlobalint2;
Virginia Polytechnic Institute and State University 15 of 77
Example of Security Check Point Queue in Extend
This example replicates the seaport example discussed in handout 7 (queuing models). The idea is to compare the simulation results with those obtained with the analytic queueing model.
The purpose of the example is to show you how simple a multiple server problem is constructed and solved using various simulation packages.
Virginia Polytechnic Institute and State University 16 of 77
Example : Level of Service at Security Checkpoints
The airport shown in the next figures has two security checkpoints for all passengers boarding aircraft. Each security check point has two x-ray machines. A survey reveals that on the average a passenger takes 45 seconds to go through the system (negative exponential distribution service time).
The arrival rate is known to be random (this equates to a Poisson distribution) with a mean arrival rate of one passenger every 25 seconds.
In the design year (2010) the demand for services is expected to grow by 60% compared to that today.
Virginia Polytechnic Institute and State University 17 of 77
Relevant Operational Questions
a) What is the current utilization of the queueing system (i.e., two x-ray machines)?
b) What should be the number of x-ray machines for the design year of this terminal (year 2010) if the maximum tolerable waiting time in the queue is 2 minutes?
c) What is the expected number of passengers at the checkpoint area on a typical day in the design year (year 2010)?
d) What is the new utilization of the future facility?
e) What is the probability that more than 4 passengers wait for service in the design year?
Virginia Polytechnic Institute and State University 18 of 77
Airport Terminal Layout
Virginia Polytechnic Institute and State University 19 of 77
Security Check Point Layout
Virginia Polytechnic Institute and State University 20 of 77
Security Check Point Model (Baseline)
Representation of the system modeled in Extend (s=2)
3V 1 2Arr. Passengers
Passengers arrive from ticket check
points
F
L W
1
Waiting Line
D
T U
Officer 1
Queue Length Trace
#
Exit(4)
732 Plotter
Security Officers
count
Leave Security
Area
D
T U
1 2 3
Rand
service time
Time in Use and Utilization Traces
Virginia Polytechnic Institute and State University 21 of 77
Extend Model with Block Labels
3V 1 2Arr. Passengers
Passengers arrive from ticket check
points
F
L W
1
Waiting Line
D
T U
Officer 1
Queue Length Trace
#
Exit(4)
732 Plotter
Security Officers
count
Leave Security
Area
D
T U
1 2 3
Rand
service time
Time in Use and Utilization Traces
Generators
QueuePlotters
Activity BlockClock
Exit Block
Virginia Polytechnic Institute and State University 22 of 77
Explanation of Extend Blocks Used
Discrete Block Generator - generates discrete entities in conjunction with discrete event models. This block has 18 different Probability Density Functions (PDF) to choose from, including an empirical
distribution format to enter real data without a PDF fit.
Queue FIFO - this structure keeps track of physical queues where the first entity arriving is the first one to be processed. Parameters L and W represent the queue length and
waiting time, respectively. Output F is a binary function that takes values of 1 when the queue is full (zero otherwise)
3V 1 2Arr. Passengers
Passengers arrive from ticket check
points
F
L W
1
Uni-queue Line
Virginia Polytechnic Institute and State University 23 of 77
Explanation of Extend Blocks Used
Activity Delay - holds an entity for a specified amount of time. Input parameter is D, the delay inside the activity block. Output parameters are T and U which represent the time in use and the utilization of the block,
respectively.
Random Number - generates a random number according to a pre-specified distribution. Input parameter are 1,2, and 3 which represent 3 input arguments used to define user-defined distributions. The single
output parameter is a random number. This block is typically used in conjunction with activity delays.
D
T U
Officer 1
1 2 3
Rand
service time
Virginia Polytechnic Institute and State University 24 of 77
Explanation of Extend Blocks Used
Exit Block (4) - destroys up to 4 items from further consideration in the simulation. The total number of entities absorbed by the block are reported. The output # connector reports the number of entities exiting the
simulation.
Plotter, Discrete Event - This block plots up to 4 parameters in the same figure. Note four input parameters in the left hand side of the block.
#
Exit(4)
473
Leave Security
Area
Queue Length Trace
Plotter
Virginia Polytechnic Institute and State University 25 of 77
Explanation of Extend Blocks Used
Executive Block - controls the duration of the simulation. This can be done either by count (i.e., a prespecified number of units or by time duration).
Note: this block has to be present in all discrete simulation models in Extend. Its location must be at the left most position in the simulation model (a necessary convention in Extend).
count
Virginia Polytechnic Institute and State University 26 of 77
Sample Results for Two-Server System (Baseline Conditions)
The results below show the utilization and waiting time instances for the security checkpoint area
0 100 200 300 400 500 6000
0.4499084
0.8998168
1.349725
1.799634
2.249542
2.69945
3.149359
3.599267
4.049176
4.499084
4.948993
5.398901
Time
Time in Use (min)Plotter, Discrete Event
0
0.07840131
0.1568026
0.2352039
0.3136052
0.3920065
0.4704079
0.5488092
0.6272105
0.7056118
0.7840131
0.8624144
0.9408157
Utilization
Waiting Time S1 Y2 Utilization S1 Wait Time S2 Y2 Utilization S2
Virginia Polytechnic Institute and State University 27 of 77
Uni-Queue Length Trace
The following diagram depicts the uni-queue length
0 100 200 300 400 500 6000
119.3333
238.6667
358
477.3333
596.6667
716
835.3333
954.6667
1074
1193.333
1312.667
1432
Time (minutes)
Leaving Security AreaSecurity Area
0
1.666667
3.333333
5
6.666667
8.333333
10
11.66667
13.33333
15
16.66667
18.33333
20
In line
11
1
11
1
11
1
11
11
11
1 Leaving Y2 In Queue Line
Virginia Polytechnic Institute and State University 28 of 77
Uni-Queue Waiting Time Trace
The following diagram depicts the waiting times vs time
0 100 200 300 400 500 6000
0.8333333
1.666667
2.5
3.333333
4.166667
5
5.833333
6.666667
7.5
8.333333
9.166667
10
Time
Waiting Time (minutes)Queue Waiting Time Trace
Waiting Time
Virginia Polytechnic Institute and State University 29 of 77
Distribution of Interarrival Times
The following plot of interarrival times was generated using Extend in the security check point model
0.000102552 0.4863909 0.9726792 1.458968 1.945256 2.431544 2.9178320
1.232143
2.464286
3.696429
4.928571
6.160714
7.392857
8.625
9.857143
11.08929
12.32143
13.55357
14.78571
Interarrival Time (minutes)
Percent ItemsItems' Distribution
% Items
Virginia Polytechnic Institute and State University 30 of 77
Discussion of Results (Baseline Year)
The following table shows typical results for the baseline year and compares them with those of the analytic model
Table 1. Results for Security Check Point System.
Parameter Analytic Queueing Model (s = 2)
Simulation Model(2 servers)
Utilization ( ) 0.900 0.916 / 0.886 a
a.Two values are reported because Extend keeps independent statistics for each server.
Expected Queue Length ( ) - per
7.60 6.96
Exp. Waiting Time in Queue ( ) - min
3.20 2.85
ρ
Lq
W q
Virginia Polytechnic Institute and State University 31 of 77
Simulation Length and Stability of Results
Simulations results require careful interpretation to guarantee stable solutions as exemplified below
0.2745101 33.56209 66.84967 100.1373 133.4248 166.7124 2000
0.3728427
0.7456853
1.118528
1.491371
1.864213
2.237056
2.609899
2.982741
3.355584
3.728427
4.101269
4.474112
Time
Time in Use (min)Plotter, Discrete Event
0
0.08134968
0.1626994
0.244049
0.3253987
0.4067484
0.4880981
0.5694477
0.6507974
0.7321471
0.8134968
0.8948464
0.9761961
Utilization
Steady-StateTransientRegion Region
Virginia Polytechnic Institute and State University 32 of 77
Part (a) Baseline Utilization Results
• From the previous graphs is evident that the utilization of each server in the baseline year is around 90%
• This result is consistent with that obtained with the analytic queueing model (i.e., multi-server and infinite source)
Virginia Polytechnic Institute and State University 33 of 77
Part (b) Horizon Year Operation
• Recalling from the analytic model that 4 x-ray machines were needed to provide levels of service below 2 minutes (in the queue)
• The corresponding Extend model and results and illustrated in the following pages
• Clearly the results of the simulation model are in close agreement with those of the analytic model presented in the queueing models section
• As a matter of simulation practice you should always check the results of your simulation models with close form solutions before developing large and more complex models
Virginia Polytechnic Institute and State University 34 of 77
Horizon Year Extend Model (4 servers)
3V 1 2Arr. Passengers
Passengers arrive from ticket check
points
F
L W
0
Uni-queue Line
D
T U
Officer 1
Queue Length Trace
#
Exit(4)
1547 Plotter
Security Officers
count
Leave Security
Area
D
T U
Officer 2
1 2 3
Rand
service time
Utilization Traces
Waiting Time Trace
D
T U
D
T U
Virginia Polytechnic Institute and State University 35 of 77
Horizon Year Model Utilization
The average utilization is reduced with 4 servers and 60% increase in the demand functions (400 minute simulation)
0.2212978 66.85108 133.4809 200.1106 266.7404 333.3702 4000
0.07387553
0.1477511
0.2216266
0.2955021
0.3693777
0.4432532
0.5171287
0.5910043
0.6648798
0.7387553
0.8126309
0.8865064
Time
UtilizationPlotter, Discrete Event
Virginia Polytechnic Institute and State University 36 of 77
Horizon Year Results ( )
The following plot shows the behavior of the waiting time function during a 400 minute simulation
W q
0 66.66667 133.3333 200 266.6667 333.3333 4000
0.25
0.5
0.75
1
1.25
1.5
1.75
2
2.25
2.5
2.75
3
Time
Waiting Time (minutes)Queue Waiting Time Trace
Waiting Time
Virginia Polytechnic Institute and State University 37 of 77
Horizon Year Queueing Characteristics
Four servers are capable of coping with 60% increase in demand (400 minute simulation)
0 66.66667 133.3333 200 266.6667 333.3333 4000
128.9167
257.8333
386.75
515.6667
644.5833
773.5
902.4167
1031.333
1160.25
1289.167
1418.083
1547
Time (minutes)
Leaving Security AreaSecurity Area
0
1.25
2.5
3.75
5
6.25
7.5
8.75
10
11.25
12.5
13.75
15
In line
11
1
11
11
1
1
11
11
11
1 Leaving Y2 In Queue Line
Virginia Polytechnic Institute and State University 38 of 77
Discussion of Results (Horizon Year)
The following table shows typical results for the horizon year and compares them with those of the analytic model
Table 2. Results for Security Check Point System.
Parameter Analytic Queueing Model (4 servers)
Horizon Year(4 servers)
Utilization ( ) 0.72 0.82 / 0.75 / 0.66 / 0.60 a
average = 0.71
a. Four values are reported because Extend keeps independent statistics for each server.
Expected Queue Length ( ) - per
1.18 0.97
Exp. Waiting Time in Queue ( ) - min
0.30 0.26
ρ
Lq
W q
Virginia Polytechnic Institute and State University 39 of 77
Answers to Parts (b) - (d)
b) The number of x-ray machines should be 4 to provide a level of service below 2 minutes
c) The the expected number of passengers at the checkpoint area on typical day in the design year should be around 0.97 persons (quite small for this type of facility but the result of the 2-min waiting time limit)
d) The utilization of the future facility should be around 71% (average of four servers)
Virginia Polytechnic Institute and State University 40 of 77
Answer to Part (e)
The Queueing Length ( ) versus time plot provides
insight on this. The red area corresponds to > 4.
Lq
0 66.66667 133.3333 200 266.6667 333.3333 4000
128.9167
257.8333
386.75
515.6667
644.5833
773.5
902.4167
1031.333
1160.25
1289.167
1418.083
1547
Time (minutes)
Leaving Security AreaSecurity Area
0
1.25
2.5
3.75
5
6.25
7.5
8.75
10
11.25
12.5
13.75
15
In line
1 Leaving Y2 In Queue Line
Lq
Virginia Polytechnic Institute and State University 41 of 77
Answer to Part (e) - continuation
Isolating a small region of the queue length function we can see more clearly the function to be integrated
82.86402 83.93834 85.01266 86.08699 87.16131 88.23563 89.30995187.7423
235.4808
283.2192
330.9577
378.6961
426.4345
474.173
521.9114
569.6498
617.3883
665.1267
712.8651
760.6036
Time (minutes)
Leaving Security AreaSecurity Area
0.07905244
0.5421237
1.005195
1.468266
1.931337
2.394409
2.85748
3.320551
3.783622
4.246694
4.709765
5.172836
5.635907
In line
1 Leaving Y2 In Queue Line
Area of Interest
Lq > 4
Virginia Polytechnic Institute and State University 42 of 77
Answer to Part (e) - continuation
• The diagram is useful to guesstimate the probability of > 4. However if we need and accurate answer we
should extract the numerical values of from the Extend output
• An equivalent way to execute this analysis is to define a new busy function in Extend such that,
= (1)
then integrate this function over time to obtain the number
the percent of instances where > 4.
Lq
Lq
B t( ) 1 if system if Lq 4>0 if system if Lq 4≤
Lq
Virginia Polytechnic Institute and State University 43 of 77
Computation of in Extend
A decision block and an integrator compute the new busy
function, , defined before in Equation (1)
Lq 4>
B t( )
3V 1 2Arr. Passengers
points
F
L W
0
Uni-queue Line
D
T U
Officer 1
Officers
Security
D
T U
Officer 2
1 2 3
Rand
service time
Utilization Traces
N
B
A
Y a>b
SR
Waiting Time Trace
D
T U
D
T U
Integrator Block
Decision Block
Plot of theIntegral of B(t)
Virginia Polytechnic Institute and State University 44 of 77
Graphical Result of the Integral of
The result below shows the integral of over time
B t( )
B t( )
0 75 150 225 3000
2.304406
4.608813
6.913219
9.217626
11.52203
13.82644
16.13085
18.43525
Time
Integral of B(t)Plotter, Discrete Event
Solid Blue GrayPat Red GrayPat Green ltGrayPat Black
Virginia Polytechnic Institute and State University 45 of 77
Avoiding Clutter with Hierarchical Blocks
• Hierarchical blocks are complex structures that contain a series of related Extend blocks
• Their main purpose is to avoid clutter in a complex model and, at the same time, provide a quick way to abstract complex behavior with minimum effort
• In simple terms, hierarchical blocks provide a way to create templates of behavior through the combination of multiple Extend blocks
• An example of a hierarchical block from a prototype landside simulation model developed at Virginia Tech is shown on the next page
Virginia Polytechnic Institute and State University 46 of 77
Hierarchical Blocks of a Simple Airport Terminal Model (ALPS - Airport Landside Performance
System)
These are some of the hierarchical blocks available in ALPS (an Extend library developed by Kulkarni and Trani, 1994)
count
Immigration
a
Heavy Acft Gate
Heavy Acft Gate
Heavy Acft Gate
Heavy Acft Gate
Heavy Acft Gate
CustomsCUSTOMSBaggage Claim
Entrance to Landside facilities
Help
Hierarchical Blocks
Virginia Polytechnic Institute and State University 47 of 77
Taxonomy of a Hierarchical Block
Hierarchical blocks contain one of more common Extend blocks as shown below
CD L W
F U
0
A
∆
Get A
b?
a
select 1 2 3
Rand
b?
a
select
CD L W
F U
0
1 2 3
Rand
CD L W
F U
0
1 2 3
Rand
F U
0
apaxIn
CD L W
F U
0
1 2 3
Rand
F
L W
0
Help
These two blocks represent the circulation in the facility
Queueing Block
Collects attributes of all passengers and sends a message to the following blocks
These blocks send the passengers to the respective baggage carousels for clooection of baggage
Baggage ClaimEquivalent to
Virginia Polytechnic Institute and State University 48 of 77
The Importance of Attributes in Simulation
• Attributes: characteristics of entities used to differentiate behavior in the simulation process (i.e.,transfer vs terminating passengers)
• Attributes move with the entity and thus can be “retrieved” in simulation blocks to perform various processes based on the attribute of an entity at time t
• Practically all simulation languages support attributes
• For example, suppose we are simulating domestic passenger baggage claim procedures. Some passengers carry carry-on luggage, others carry full baggage that needs to be retrieved from a direct feed baggage system.
Virginia Polytechnic Institute and State University 49 of 77
Attributes (Illustration)
• If the percentages of passengers carrying full bags is known a random number generator and an attribute block can be used to model both types of passengers
• In this example an attributes block is used to assign either carry-on or full bags to each passenger
start
VF
L W
550
D
T U
CD L W
F U
10 paxoutA
Set A
Help
Program Block
Attributes Block
Queueing Block
Activity Delay Block Animation
BlockMultiple Activity Block
Virginia Polytechnic Institute and State University 50 of 77
APM Simulation Model (Lin and Trani, 1995)
Purpose of research model
• To model individual passenger and APM vehicle movement at airports
• To determine the sensitivity of system performance
• To estimate the APM vehicle energy consumption
• To examine the flexibility of an APM system
Generically we call this model APMSIM
Virginia Polytechnic Institute and State University 51 of 77
APM Requirements Analysis
Level of Service Analysis
APM Demand Analysis
Capacity Analysis
Flow Analysis
Energy Consumption Analysis
Virginia Polytechnic Institute and State University 52 of 77
Methodology
Developed a hybrid simulation model in ExtendTM
time
Process
Event 1 Event 2 Event 3 Event 4 Event 5entityarrival
servicebegun ontask 1
servicebegun ontask 2
serviceended ontask 1
serviceended ontask 2
Activity 1
Activity 2
Virginia Polytechnic Institute and State University 53 of 77
APM Simulation Model Description
• APM Simulation Model
• APM station model
• APM guideway model
• Energy consumption model
• Simulation Model Logic
• Passenger flows
• Vehicular flows
Virginia Polytechnic Institute and State University 54 of 77
Sample Schematic of Station Model
Veh02_In
Veh12_Out
Veh02_Out
Veh12_In
Veh11_Out
Veh01_In
Veh11_In
Veh01_OutD B1
B1D B2
B2Transit Unit
D B1
B1D B2
B2Transit Unit
EscalatorB
D
B
D
EscalatorB
D
B
D
StairwayD
BB
D
ElevatorB
D
B
D
D
B
D
B
CirculationPassageway
D B1
D B1 B2
B2
D
B
Platform
(1) Pax.Schedule
Info.
Boarding Passenger Flow Direction
Deboarding Passenger Flow Direction
Virginia Polytechnic Institute and State University 55 of 77
Sample APM Guideway Model
Double-Lane Loop
Pinched Loop with Turnbacks
A
B
A
B
B
A
B
A
StationStation
StationStation
BA
BA
StationStation Station
Virginia Polytechnic Institute and State University 56 of 77
APM Energy Consumption Model
The purpose of this submodel is to estimate the following:
• Speed
• Acceleration
• Travel Time
• Travel Distance
• Power Requirements
• Energy Consumption
• LOS (Level of Service)
• Occupancy and Load Factor
Virginia Polytechnic Institute and State University 57 of 77
Energy Modeling
• David’s Equation, Power required and energy consumed
• Integrate these relationships over time (use Extend’s RK-4 integration algorithms
R (a+ r) = K0 +K1
w+ B(V) +
CAV2
wn
E = 3.6 ×10 -6 Pdt0
t
∫
P =T(V)
Virginia Polytechnic Institute and State University 58 of 77
Typical Station Results
• Travel time for an individual passenger
• Average waiting time for a facility or a TU
• Total number of passengers arriving at or leaving a station or a facility
• Queue length at a facility
• LOS of a facility in terms of area per passenger
Virginia Polytechnic Institute and State University 59 of 77
Typical Guideway Results
For a single guideway section or for the complete guideway network the model estimates:
• Travel time of an individual TU or passenger
• Occupancy and load factors of a TU
• TU speed, acceleration, deceleration, and travel distance
• TU power requirements and energy consumption
• Number of TUs in a guideway or a system
• LOS in a vehicle
Virginia Polytechnic Institute and State University 60 of 77
Sample Application of the APMSIM Model
We modeled Atlanta Hartsfield Intl. Airport in the US
Concourse A
InternationalConcourse
Concourse B Concourse C Concourse D
Ticketing
APM Station
Station Doors
305 m 305 m305 m278 m
NorthGuideway
SouthGuideway
52 m 76 m 52 m
41 m
88 m 130 m 346 m
305 m
305 m 305 m
52 m305 m
305 m
305 m221 m Bypass
Baggage
368 m
Concourse E
242 m 69 m 38 m
Maintenance
305 m
Virginia Polytechnic Institute and State University 61 of 77
Atlanta APMSIM Guideway Model
ABC
count
BAC
BA C A
B
C
BA
BA
Station
Station
Station
Station
Station
Station
BA
VehicleSchedule
Bypass
Bypass
Station
Turnaround
Baggage Station Ticketing Station Concourse A
Concourse B
Concourse C Concourse D Concourse E
North Guideway
South Guideway
South Guideway
North Guideway
North Guideway
South Guideway
Virginia Polytechnic Institute and State University 62 of 77
Atlanta APMSIM Station Results
Passengers entering and leaving a station
600 666.6667 733.3333 800 866.6667 933.3333 10000
32
64
96
128
160
192
224
256
288
320
352
384
Simulation Time (seconds)
Total No. of PAX (passengers)PAX Entering/Leaving a Platfrom
Pax Entering Pax Leaving
Virginia Polytechnic Institute and State University 63 of 77
ATL APMSIM Simulation Results
Passengers waiting at a station (on the APM platform area)
600 666.6667 733.3333 800 866.6667 933.3333 10000
10
20
30
40
50
60
70
80
90
100
110
120
Simulation Time (seconds)
No. of Passengers (passengers)No of Passengers on the Platform
Virginia Polytechnic Institute and State University 64 of 77
ATL APMSIM Results
Level of service at the APM platform
600 666.6667 733.3333 800 866.6667 933.3333 10000
2.5
5
7.5
10
12.5
15
17.5
20
22.5
25
27.5
30
Simulation Time (seconds)
Area per Passenger (m^2/pax)LOS on the Platform
LOS Equivalent LOS
Virginia Polytechnic Institute and State University 65 of 77
ATL APMSIM Simulation Results
Number of Passengers Entering a Concourse
600 666.6667 733.3333 800 866.6667 933.3333 10000
20.83333
41.66667
62.5
83.33333
104.1667
125
145.8333
166.6667
187.5
208.3333
229.1667
250
Simulation Time (seconds)
Accum. PAX Flow (passengers)Passengers Entering the Concours
0
0.75
1.5
2.25
3
3.75
4.5
5.25
6
6.75
7.5
8.25
9
PAX Flow
Accum. Pax Flow Y2 Pax Flow
Virginia Polytechnic Institute and State University 66 of 77
ATL APMSIM Development Conclusions
These are conclusions of this model development effort:
• Able to model various types of APM system configurations
• Useful to estimate the effects of changes in the system by modifying the input data
• Easy to model alternative APM service concepts
• Capable of simulating an APM system for different scenarios
Virginia Polytechnic Institute and State University 67 of 77
APM or Other Transport Modes in Terminal Airport Models
• As demonstrated in the APMSIM model average statistics about LOS can be obtained in discrete and hybrid simulations
• These statistics can, in the end, be used to compare the static and macroscopic values typically used in airport terminal design
• As engineers and analysts we have to get the point to decision makers that airport terminals cannot be viewed and designed using static metrics
• It is more relevant to know how queues form and dissipate at processing facilities than computing a “perfect-gas molecular” equivalent (area per passenger)
Virginia Polytechnic Institute and State University 68 of 77
References
1) Lin,Y. A Simulation Model of an Automated People Mover at Airports. M.S. Thesis. Virginia Polytechnic Institute and State University, Blacksburg, VA 24061.
2) Kulkarni, M. Development of a Landside Terminal Simulation Model. M.S. Thesis. Virginia Polytechnic Institute and State University, Blacksburg, VA 24061.
3) Fruin, J.J. Designing for Pedestrians. in Public Transportation Systems. Hoel and Gray: Editors. John Wiley and Sons, New York, 1993.
4) IATA. Airport Development Reference Manual: 8th Edition. International Airline Transport Association, Montreal, 1995.
Virginia Polytechnic Institute and State University 69 of 77
Virginia Polytechnic Institute and State University 70 of 77
Virginia Polytechnic Institute and State University 71 of 77
Virginia Polytechnic Institute and State University 72 of 77
Virginia Polytechnic Institute and State University 73 of 77
Virginia Polytechnic Institute and State University 74 of 77
Virginia Polytechnic Institute and State University 75 of 77
Virginia Polytechnic Institute and State University 76 of 77
Virginia Polytechnic Institute and State University 77 of 77