development of methods of increasing terminal flexibility ... · 30/09/2016 · development of...
Post on 22-Jul-2020
6 Views
Preview:
TRANSCRIPT
Architecture
Specialists in Computer Architecture
TechnologyCorporation
Architecture
Specialists in Computer Architecture
TechnologyCorporation
Development of Methods of Increasing
Terminal Flexibility and Control
Authority
NASA Contract: NNA14AC42C
Option Year 1 Final Presentation
September 30, 2016
Version #1
Architecture Technology Corporation
1Final Briefing, 30 Sep 2016, Version 1
Outline
• Background and Overview of Departure Management What-if
Analysis
• Option Year 1 Objectives and Accomplishments
• What-if Analysis Tool Enhancements
• Traffic and Weather Scenarios
• What-if Analysis Tool Verification
• What-if Evaluation Metrics
• What-if Evaluation Example
• What-if Analysis Tool Demonstration
• Recommendations
2Final Briefing, 30 Sep 2016, Version 1
What-if Analysis Tool
Background
• NASA Airspace Technology Demonstration 2 (ATD-2)
– Integrated arrival-departure-surface traffic management tools and
operations enable ideal trajectories for departures
– Delay at gate, unimpeded taxi on the airport surface, minimum time in
departure runway queue, and continuous climb to cruise altitude
• FAA Surface Collaborative Decision Making (CDM) Concept of
Operations
– Departure Management Programs (DMPs) to provide strategic Target
Movement Area Entry Times (TMATs) to control surface traffic levels
• Interfaces of NASA ATD-2 with FAA Surface CDM
– Interface between Spot and Runway Departure Advisor (SARDA) runway
takeoff and spot release sequence & schedule and DMP TMATs
– Ramp control to meet strategic TMATs
• What-if analysis
– Strategic planning of DMPs to mitigate effects of demand/capacity
imbalances at airport under forecast operating conditions3Final Briefing, 30 Sep 2016, Version 1
Departure Metering What-if
Analysis Concept Overview
4Final Briefing, 30 Sep 2016, Version 1
I ----------------,
: Far-term Time Horizon • I
: • Traffic Schedule • I
: • Operating Conditions I 1 • Departure Restrictions l---------- ----------_-----------
Assess impact,::;\ on queue \.!I lengths, taxi time, gate delay, etc.
Predict per runway need for DMP; DMP start time and duration
What-if Analysis Tool
,------- --------: Demand/Capacity : : Imbalance, Initial : I I , DMP Parameters , I I --------
Departure Reservoir Coordinator (DRC}
Analyze impact of potential changes to restrictions, runway rates, and scheduled traffic
DRC implements
Evaluate effect of DMP parameters on managing demand/capacity imbalances and meeting airport performance objectives
© DMPduring
r - - - - - - - - - - - - I using selected : Selected DMP I parameters I ;'-----1----------_j • Parameters : l ____________ ,
Departure Metering System
I ----- -------1 : Target Movement 1
I I
, Area Entry Times 1 I I
(TMATs} : -------•
ATD-2 Traffic Management
System
Actual Traffic Movements
I I ----------------
<tJ-aoEING
~
I ---------------1 : Near-term Time Horizon : • Traffic schedule : • Operating conditions I , • Departure restrictions , l - - - - - - - - - - - - - - - - - - - _I
DMP specifies TMATs for departures as per near-term conditions
I -----------------,
: Current Time , I
• Traffic situation : • Operating conditions I , • Departure restrictions : 1 ____________________ 1
ATD-2 system performs integrated arrival-departure management
Option Year 1 Objectives
SOW
• Develop and refine the what-if capability
– Airport surface
– Terminal airspace
– Metrics and interfaces
– Traffic and weather scenarios
– DMP parameters and scope
• Use what-if analysis capability
– Specify DMP parameters for CLT under different traffic and weather
conditions impacting departure traffic flow
– Evaluate effectiveness of DMPs in mitigating impacts of traffic flow
inefficiencies
• E.g., reducing delays during surface taxi & airborne transit
5Final Briefing, 30 Sep 2016, Version 1
Option Year 1
Accomplishments
6Final Briefing, 30 Sep 2016, Version 1
Objective Accomplishments
Airport surface • Investigated and documented CLT surface operations
• Implemented node-link modeling of airport runways, spots and terminal gates for different
configurations
• Specified modeling parameters from CLT operations data and references, Base Year analyses
• Verified implementation of models, compared simulation results to FAA ASPM
• Developed detailed models of traffic flow interactions of runways, taxiways, non-movement area
Terminal
airspace
• Implemented and verified modeling of time period miles-in-trail restrictions for departure fixes
• Implemented and verified miles-in-trail restrictions for departure runways to meet fix restrictions
• Modeled link transit times from Base Year high-fidelity departure simulation data
Metrics &
interfaces
• Implemented metrics and interfaces for DRC to assess airport departure and arrival traffic flow
and design DMPs
• Implemented interfaces to configure and conduct What-if analysis and assess results
DMP
parameters &
scope
• Summarized specifications for DMPs from FAA Surface Collaborative Decision Making ConOps
• Implemented explicit control of Target Departure Queue Length for individual departure runways
• Implemented methods to accommodate multiple flow restrictions on departures
• Implemented automatic DMP start & end times from runway queue data
Traffic &
weather
scenarios
• Identified recent operational days for idealized traffic schedule and traffic “disturbance”
scenarios
• Created input files for What-if Tool from traffic and restrictions data for those days
DMP
Evaluations
• Used What-if analysis tool to conduct demand analysis and DMP implementation for different
historical and notional traffic & weather scenarios
• Documented results & developed demonstrations
What-if Analysis Tool
Capabilities
• Adapt to changes in airport and airspace operating conditions
– Runway configurations & rates
– Traffic levels & airport/airspace resource utilizations
– Departure fix restrictions
• Design & emulate Departure Management Program
– Scheduling of gate pushback times & TMATs
– Meter flights to control runway queue length
– Account for per-runway departure rates & multiple departure restrictions
• Rapidly evaluate airport traffic performance
– Evaluate variety of operating and DMP alternatives and uncertainty
– DMP go-no go screening, start & end times, particular runways
• Present key departure and arrival performance metrics for detailed
and aggregate performance assessment
– Metrics important to airport operations
– Nature of demand characteristics and airport traffic response
– Time period- and runway-based assessment for detailed understanding
– Aggregate assessment for comprehensive overview
7Final Briefing, 30 Sep 2016, Version 1
CLT Airport Modeling
Enhancements• Node-link modeling of airport and
airspace
– Synthesized from gate/spot/runway/fix
routes of flights in traffic file
– Provides modeling flexibility to balance
fidelity with simulation time
• Nodes
– Queueing points where congestion
occurs
– Nodes
• Gates, spots, runways, arrival and
departure fixes abstracted as nodes
– Parameters
• Aircraft service time, Ts
• Maximum queue size, Qmax
– First in first out at specified service rate
– Traffic in-flow versus out-flow
determines flight delay
• Links
– Transit segments between nodes
– Parameter
• Transit Time TT
8Final Briefing, 30 Sep 2016, Version 1
Aircraft undergoing transit
Transit time TT
GATE SPOT RWY FIX
Link Model
Entities modeled as nodes
Aircraft served
at rate Ts
Aircraft
waiting
in queue
Empty
queue
slots
Max queue size QMAX (if any)
NODE
Elements of node
Node Model
Arbitrary route system geometries
may be modeled
<tJ-aoEING
~
1· + +
12N
12
10
82
1
18C
18L
A Gates
D Gates
E Gates
C Gates
B Gates
CLT Airport Surface
Departure Modeling Example
9Final Briefing, 30 Sep 2016, Version 1
Gate nodes
Spot node
Runway node
Gate-spot link
Spot-runway link
Node-link modeling of gate-spot-runway routes
• Nodes: traffic
interaction &
rate limits
• Links:
undelayed
transit
Terminal Charts - PARKING GATES n :r ~ W80-57.J ll'80-57.1 G'> ; ~
Nll •l l .S , i ; • 'll i N m N3S- IJ.4 .. ., • 3 0 < . ~
NlS- IU
N
Nl5 IJ 2
V @
Ill ::: $ ,z NlS-1~.1
i ~ u 0
~
~ ,. i:i NlS-IJ,O :r .: :!l V
~ W80-57.3 11'80- 57.2 !>
W80-57.0 WS0-56.9 11'80-56.e WS0-56.7 W80-5M
Ramp Conlrol
North 130.07 South 131.6 East 129.22
I© DEPARTURE CALL SPOTS 11 1-.. RAMP CONTROL SECTORS!
Weit 129.8
Dual Taxi betwHn Departure Call Spot i 11/ 12 and 13N / 13S m tr lcted to aircraft with wi ngspans less than 171' (52m) .
Dua l Taxi between Departure Ca ll Spot: 22 /23 and 24N /24S re:tricted to aircraft with wi ngspans less than 118' (36m).
• CONTROL TOWER
<tJ-aoEING
~
11'80- 56.2 \1'80-56.1
('l I
?i; Ci r- :I: o • =1 ;:o m r-'- O CJ-•
D o_,. C m G) , r-)>z 1/)
... Ci ~)>, r- ;:o
18C
18L
CLT Terminal Airspace
Departure Modeling Example
10Final Briefing, 30 Sep 2016, Version 1
Runway node
Departure fix node
Runway-fix link
• Nodes: traffic
interaction &
rate limits
• Links:
undelayed
transit
Node-link modeling of runway-fix routes
<tJ-aoEING
~
ANDYS
CLT Node & Link Parameters/Attributes/Behavior
11Final Briefing, 30 Sep 2016, Version 1
Node
type
Service time,
Minutes
Size limit,
aircraft
Data source
Gate 30.0 1 Estimation
Spot 1.0 N/A Estimation
Runway 2.0 N/A 30 aircraft/hour
Fix 1.7 1 SME-specified 7 miles @ 250
knots
Link type Transit time,
minutes
Data source
Gate-spot 4.0 SOSS simulation of CLT
Spot-runway 2.0 SOSS simulation of CLT
Runway-fix 14.0 Flight simulations of MERIL departures
• Service time: Minimum time to process aircraft, models rate limit of traffic passing through
node, e.g., time interval of runway departure rate, in-trail spacing of departures crossing fix
• Size limit: Number of aircraft that can be waiting for service, e.g., number of departures that
taxiway can fit, number of aircraft that can occupy terminal gate
• Transit time: Undelayed transit time between nodes
<tJ-aoEING
CLT Specialized Node Models
• Runway nodes
– Node exit time for departures as per
• Node service time
• Service time for miles-in-trail restrictions at departure fix
• Gate nodes
– Departure entry/exit times fixed
• Entry at scheduled gate entry time
• Exit at gate occupancy time or DMP-scheduled pushback time
– Arrivals delayed entry to gate until occupancy time window is
available
12Final Briefing, 30 Sep 2016, Version 1
Fix Miles In Trail Time Start, Min Time End, Min
MERIL 20 50 150
BUCKL 15 75 200
<tJ-aoEING
What-if Tool Enhancements
DMP EmulationKey Features
• Generates TMATs to
absorb delay at the gate
• Satisfies multiple
constraints
– Minimum Gate Occupancy
Spacing
– Minimum Departure Spacing
at Runways
– Departure Fix Flow
Restrictions
• Attempts to maintain a
specified runway queue
size
13Final Briefing, 30 Sep 2016, Version 1
Gate Entry/Ex it Spacing
Gate Entry
~ Gates
Gate Ex i t_~ ~ - -, / YJ;., ; - -
Gate Ex it Spacing'~/ ,:
, , , J:lunway Spacing
--~--' ,;1!'' --~-' ,
<tJ-aoEING
D eparture Fix ,_ l::,.
, , '-::t , , ""'11 ,
, '" , M IT Departure Flow M eter ing '
D eparture Fix l::,.
M IT :~ : Departure : : Flow Meterino >i-.:
b o '
DMP Emulation
Components
• High Level Functionality
– Three main components
14Final Briefing, 30 Sep 2016, Version 1
Sort Flights by reference Points
RunwaysFlow
RestrictionsGates
Sort and Space Flights by time at each
reference point
Iterative Spacing Routine
Initial Attempt at full schedule
solution
‘Fine-tunes’ schedule through iteration until all constraints
are satisfied
RemainingConflicts
Component 1 Component 2 Component 3
<tJ-aoEING
DMP Emulation
Process Steps
• Step 1: Sort by Reference Point
• Step 2: Initial Sorting and Spacing
• Step 3: Iterative solution for
convergence
15Final Briefing, 30 Sep 2016, Version 1
Sort at Flow Restrictions
Space at Gate Entry
Space at Flow Restrictions
Sort at Gate Entry
Sort at Runways
Space at Runways
Space at Gate Exit
Sort at Gate Exit
Space at Gate Entry
Space at Flow Restrictions
Sort at Gate Entry
Space at Gate Entry
Space at Flow Restrictions
Space at Runways
Space at Gate Exit
Was
delay
added?
yes
no
Solution Convergence
Step 2
Step 3
<tJ-aoEING
1 .I,
1
"' I I
'1,-
l
I I t
I
DMP Emulation
Aircraft Spacing Algorithm
16Final Briefing, 30 Sep 2016, Version 1
1
( )
i i
space ref pt ref pt
space sp delay sp space
gate exit delay
gate rwy
runway gate exit transit delay
rwy fix
dep fix runway transit delay
t t t
if t t t t t
t t
t t t t
t t t t
Stop
Get i+1 Flight
Is flight (i+1) spacing after
(ith )sufficient?
no
yes
Get ith Flight
Advance i
Add Delay until i+1spacing behind ith
is sufficient
More Flights
?
no
yes
Start
i=1
Explicit algorithm
minimum required spacing
Existing spacing
Algorithm is the sameregardless of referencepoint
Delay is always added at the gate and cascades through all downstream reference points
3t
1t
2t
tn-1
tn
1t
2t
3t
n-1t
nt
spt
spt
spt
...
...
Sorted but not
spaced
Spaced
- - -I
I :+ : I I
I I
I I ,+ ' I I
I I
:+r I I
I I :+ : I I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I I
I I
I I
: ~· :---->: I I I
I I I
:~ : I y I
:~ : I y I
I I
._ - - - I
:+:l I
<tJ-aoEING
l '----->
DMP Emulation
Runway Queue Size Control
17Final Briefing, 30 Sep 2016, Version 1
rin rout
q =.
rin - rout
QUEUE
in outq r r dt
1 i ii i in out iq q r r t Discrete
1out
rwy
rt
1
i
in
sp
rt
q t ie q q
i
q
sp rwy q rwy
t
et t K t
q
1
( )i i
i i
spacing runway runway
spacing sp delay sp spacing
t t t
if t t t t t
1
11
ii i sp
rwy
q q tt
Aircraft queue, q, forms
when supply rate, rin,
exceeds service rate, rout
Supply & service rates rin & rout as
inter-flight times tsp & trwy
Queue size, qi, feedback control of supply
rate/spacing, tsp, to meet target queue size, qt
0iq
Aircraft supply
rate & spacingAircraft service
rate & spacing
Delay aircraft queue entry time as needed to
satisfy inter-aircraft supply spacing, tsp
Continuous
<tJ-aoEING
• ~ IL___ _I~ •
Traffic & Weather Scenarios
Development Methods
18Final Briefing, 30 Sep 2016, Version 1
• Information sources for scenario initialization
– Weather Underground for historical weather to select scenario days
– ASDI in-bound fix/TRACON entry fix position and time (ATA) data
– Out/Off/On/In (OOOI) from a major CLT airline operator for August 2014*
– NASA Restrictions May through Dec 2014 • Expected Departure Clearance Time (EDCT)
• Call for Release (CFR)
• Miles in Trail (MIT): used fix location and miles spacing values
– Ramp controller procedures to infer spot allocation
• Scenario preparation methodology
– Capture all arrivals/departures from 5 AM to midnight
– Infer tail number connectivity by associating scheduled gate IN-OUT times per gate
and aircraft type
– Start time for each departure based on IN time of associated arrival
– Spot assignments from airport layout, gate/runway pairs and airline Standard
Operating Procedures (SOP)
* Note: Boeing obtained Official Airline Guide (OAG) data for the 3rd quarter of 2014 as reported in the “Traffic and Weather Identification and Modeling
Document (Contract CDRL 4.6). However, the team relied entirely on the out/off/on/in (OOOI) schedule data to build the scenario.
Traffic & Weather Scenarios
Reference Data For Days
19Final Briefing, 30 Sep 2016, Version 1
Type Date in 2014
Weather at CLT # of EDCT
# of CFR
# of MIT restrictions
Comment
Baseline (Good Weather)
August 8
Clear, visibility of 8 NM 4 25 42 Usual restrictions for CLT
Disruptive Events (Bad Weather)
August 11
Rainfall at CLT reducesvisibility to 1 NM
1 25 45 Storm moving through CLT
August 15
VMC/VFR conditions, no flow reversals
7 42 76 Weather near Atlanta with restrictions imposed by Atlanta ARTCC
August 18
2 41 180 Heavy volume restrictions due to extreme rainfall in TN and NE Alabama
Entry Fix
Exit Fix
Note: entry/exit fix positions
are outside the diagram and
are represented notionally
Scenario Days
CLT Operational Constraints Arrival Flows & TRACON Fixes
<tJ-aoEING
~
--
Weather and Traffic Scenarios
Departure Fix Restrictions
• CLT Departure Traffic With Restrictions for 8/18/2014
– What-if tool models MIT restrictions
20Final Briefing, 30 Sep 2016, Version 1
Fix Miles In
Trail
Time
Start, Min
Time
End, Min
MERIL10 554 559
MERIL15 570 732
MERIL15 780 970
MERIL20 970 1005
MERIL25 1005 1090
MERIL25 1139 1261
MERIL10 475 505
MERIL15 1100 1185
<tJ-aoEING
Histogram of scheduled gate departure times for 8/8/2014 baseline traffic
scenario
Weather and Traffic ScenariosBaseline Scenario Surface Traffic Demand
21Final Briefing, 30 Sep 2016, Version 1
Histogram: number of scheduled
departures by ¼ hour
3-5
30
25
~ L 0
-~ 20 0
L ®
~ 15 :J z
10
5
0 200
I
I T
400
<tJ-aoEING
~
Dis1trib uti,on of OUT 1ti m,es I I I I I
-
-
-
-
-
-
~ { J I i J ~ l.n 600 800 1000 1200 1400 1600
l"ime of dary ( m i1n)i
What-if Analysis Tool
Verification• Traffic Simulation
– Parameter adherence
• Verified traffic flow adheres to link transit times, node service rates, node queue length limits,
departure fix restriction spacing & time period
– Comparison to FAA ASPM for 8 August 2014
• Input traffic schedule
– IN-OUT times differ from airline-scheduled times
• Departure throughput & taxi-out times
– Departure rates: Comparable maximum and total average, different hourly averages
– Taxi-out times: Comparable hourly and total averages
• Arrival throughput & taxi-in times
– Arrival rates: Comparable maximum, hourly and total averages
– Taxi-in times: Simulation higher due to gate occupancy and utilization modeling
• DMP Emulation
– Parameter adherence
• Verified scheduled departures adhere to runway rates, departure fix restrictions and gate
occupancy restrictions
– Traffic control
• Verify metering meets traffic performance requirements
22Final Briefing, 30 Sep 2016, Version 1
Airport Evaluation Metrics
Runway Demand & Capacity
23Final Briefing, 30 Sep 2016, Version 1
Time-bin traffic demand
for each runway
Per-runway
capacity
Arrivals
Departures
Per-runway capacity of
7-8 aircraft/quarter hour
Per-runway
capacity of 7-8
aircraft/quarter hour
0
" ,,
14 ,--------_l
12
10
5 6 z
4
r-------,---------,-------------,-----~ fiii,aR1 ~
o o~ ~~~~----;----------:-1:-----0.5 1.5 2 Simulated Time, Hours 2.5
<tJ-aoEING
14
4
o o;-____ LJ_JE__L 0.5
1.5~- o.J...J.--;--2---__J ____ _
Simulated Time, Hours 2.5 3
Airport Evaluation Metrics
Departure Performance
24Final Briefing, 30 Sep 2016, Version 1
Departure
Runway
Average
Throughput,
Departures
Per Hour
Average
Queue
Length,
Departures
Maximum
Queue
Length,
Departures
Minimum
Queue
Length,
Departures
Average Taxi
Out Time,
Minutes
Average
Gate Delay,
Minutes
18C 29 3 12 0 23.6 0
18L 26 9 25 0 41.9 0
Key performance metrics of
throughput, congestion,
taxi time and gate delay
Airport runway traffic
performance for detailed
understanding of operations
Time-bin presentation to see
trends & variability
Time period presentation to
know total performance
Limits for comparison
¢,
~ ~ <
Number of Actual Takeoffs 8 ,--------..,----r;:::r====;i
- 18C C]18L
7
5
04 a; ,::,
20
¢e 15 ~ ~ < 0 a;
,::, E ::, Z 10
5
70 Average Taxi Out Time
~ L
60
50
"' Ql "5 40 C:
~ c1i E i= ·x 30 "' I-
20
10
<tJ-aoEING
"' .l!l ::, C:
~ Average Gate Departure Delay 1 ====--------,---,
0.9
0.8
0.7
0.6
~ 0.5
i ~
0.4
0.3
0.2
0.1
Airport Evaluation Metrics
Arrival Performance
25Final Briefing, 30 Sep 2016, Version 1
Arrival
Runway
Average
Throughput,
Arrivals Per
Hour
Average Taxi
In Time,
Minutes
18C 29 23.6
18L 26 41.9
Key performance metrics of
throughput and taxi time
Airport runway traffic
performance for detailed
understanding of operations
Time-bin presentation to see
trends & variability
Time period presentation to
know total performance
Number of Actual Landings by Runway
._____ __ ______.1 z
3
2
o~~-~~-~~-~--~--~-~
0 0.5 1.5 2 2.5 3
Simulated Time, Hours
<tJ-aoEING
~ Average Taxi In Time by Runway
30 ~ ~ -~--~--~--~--~-~
1/)
2 ::, C:
~
25
20
<Ii' 15 E i=
! 10
5
~
o~~-~~-~~-~--~--~-~
0 0.5 1.5 2 2.5 3
Simulated Time, Hours
What-if A
naly
sis
Tool
Exam
ple
Evalu
atio
n
Tra
ffic s
ch
ed
ule
•A
ugu
st 8
, 2014, 2
:08 p
.m. to
5:5
8 p
.m. lo
cal
CLT
time
26
Fin
al B
riefin
g, 3
0 S
ep 2
016, V
ers
ion 1
Fix
Mile
s In
Tra
il
Tim
e
Sta
rt,M
in
Tim
e
En
d, M
in
ME
RIL
10
49
ME
RIL
15
20
182
Tim
e =
0 M
in
AN
DY
SB
UC
KL
DE
BIE
JA
CA
LL
ILL
SM
ER
ILN
AL
EY
ZA
VE
R
18
C1
27
24
33
27
18
L11
19
31
63
64
3
Dep
artu
re re
stric
tion
s
•R
estric
tions d
ue to
nearb
y ra
insto
rms
on
Au
gu
st 1
8, 2
01
4, 2
:08
p.m
. to 5
:58
p.m
. local C
LT
time
Scen
ario
Tim
e =
23
0 M
in
i!l N ~
What-if Evaluation,
Airport Demand Analysis
27Final Briefing, 30 Sep 2016, Version 1
Demand intermittently
exceeds RDR of 7-8
aircraft/quarter hour
Departure Demand Airport Departure Performance
Departures push back at
airline-scheduled times
• Runway throughput intermittently saturates
• Runway departure queues exceed target length
• Runway departures exhibit excessive taxi-out times
14~---~_N_u_m_b_e_r_o_f~D_e_p~a_rt_u_r_e_P~u_s_h_b_a_c_k_s~---~
~ 12
0 ... .2l 6 E :,
z 4
2
o ~-~~~~~~~~~~~~~~~~~~--~
0 2 3 4 5 Simulated Time, Hours
Number of Actual Takeoffs 8
7
6
5
2
o~~---~~-~ 0 2 3 4 5
Simulated Time, Hours
10
8
o 6 ai .0 E :, z
4
2
Average Queue Length
~---
0
II 2 3 4 5
Simulated Time, Hours
<tJ-aoEING
35 Average Taxi Out Time
~ 30
25
10
5
o~---~~--~ 0 2 3 4 5
Simulated Time, Hours
~
1~ verage Gate Departure Delay
~ 10
8
4
2
o~~-~-~-~~
0 2 3 4 5 Simulated Time, Hours
What-if Evaluation,
DMP Analysis
28Final Briefing, 30 Sep 2016, Version 1
Departure Demand Airport Departure Performance
Departures push back at
DMP-scheduled times
• Start: 60 min
• End: 289 min
Demand more closely
complies with RDR of 7-8
aircraft/quarter hour• Runway throughput maintained
• Runway departure queues closer to target length
• Average taxi-out times reduced
• Gate holding delay introduced
Number of Departure Pushbacks 14~--~ -~ ---~----~---~---~
~ 12
0 ... i 6 E :,
z 4
I I I I
.J -I I
I I I
~ ---,-. ·~----1 .
Simulated Time, Hours
Number of Actual Takeoffs 8
7
6
5
~ ~ < 04 ai .0 E :, z
3
2
0 0
" - 18C C]18L
2 3 4 5 Simulated Time, Hours
12 Average Queue Length
~ L
10
8
<I= ~ ~ < 0 6 ai .0 E :, z
4
2
2 3 4 5 Simulated Time, Hours
<tJ-aoEING
~
Average Taxi Out Time Average Gate Departure Delay 35 12
~ L ~ L
30 10
25
8
(/J Cl)
(/J :i 20 C 2 ~
:, C
ai ~ 6 E ;>"; i= a,
·x 15 <ii 0
~
4
10
2 5
0 0 I d 0 2 3 4 5 0 2 3 4 5
Simulated Time, Hours Simulated Time, Hours
What-if Evaluation
Comparison, Departures
29Final Briefing, 30 Sep 2016, Version 1
Avg Throughput,Departures/Hour
Avg QueueLength,
Departures
Max QueueLength,
Departures
Avg Taxi Out Time,Minutes
Avg Gate Delay,Minutes
Without Departure Metering 22 3 12 17.1 0
With Departure Metering 22 1 4 12.6 2.9
0
5
10
15
20
25
CLT Departure Runway 18L • Departure metering program
effective
– Runway departure throughput
maintained
– Runway queue lengths reduced
• More significant reductions
for runway 18L departures
– Arrivals to runway 18C
interfering with planned runway
departure rate
• Average taxi-out times
reduced
– More significant reduction of 4.5
minutes for departures from
runway 18L
– Arrivals to 18C impacting
departure taxi-out delay
• •
25
20
15
10
5
0
• Without Departure Metering
• With Departure Metering
CLT Departure Runway 18C
Avg Throughput,
De pa rtu res/ Hour
20
20
Avg Queue Length,
Departures
4
3
Max Queue Length,
Departures
14
10
Avg Tax i Out Time, Avg Gate Delay,
Minutes Minutes
18.6 0
16.3 3.0
<tJ-aoEING
~
What-if Evaluation
Comparison, Arrivals
30Final Briefing, 30 Sep 2016, Version 1
avgThroughput,Arrivals/Hour
avg Taxi InTime, Minutes
Without Departure Metering 20 32.7
With Departure Metering 20 34.3
05
10152025303540
CLT Arrival Runway 18R
avgThroughput,Arrivals/Hour
avg Taxi InTime, Minutes
Without Departure Metering 17 33.8
With Departure Metering 17 34.8
05
10152025303540
CLT Arrival Runway 23
avgThroughput,Arrivals/Hour
avg Taxi InTime, Minutes
Without Departure Metering 7 25.2
With Departure Metering 7 27.3
05
10152025303540
CLT Arrival Runway 18C
• Arrival taxi-in time increases due to
departure gate holding
– Average taxi-in delay increase due to
increased gate occupancy of departures
<tJ-aoEING
• •
Tool Demonstration
• Test case of delaying DMP start time
– Traffic schedule
• hitl6-training-advisory.list_data
• CLT south flow
• Departures: 52 from 18L, 37 from 18C
• Arrivals: 41 from 18R, 38 from 23
– Departure restrictions
• TrafficFlowRestrictions.csv
– Departure management program
• Start time, min: 60, 30
31Final Briefing, 30 Sep 2016, Version 1
Fix Miles In Trail Time Start, Min Time End, Min
MERIL 20 50 150
BUCKL 15 75 200
<tJ-aoEING
35---,--------."-P_..u,.,.s._,h.,.ba~c;;,k':s;_;&~L~al"lncl_di_riin~gii;s~-~ ===~ • Departures • Arrivals
30
¢: 25 ~ ~ ~20 0 215 E :, z
10
5
g ~ Time, Minutes
0
"'
Summary
• Developed and demonstrated prototype What-if Analysis Tool for
strategic assessment of airport traffic and planning of DMPs/TMATs
– Adaptable to forecast airport operating conditions
– Rapid evaluation of traffic
– Emulation of DMP
– Metrics & presentation to understand traffic behavior & assess airport
performance
– Supports exploring airport traffic behavior & DMP implementation
• Applied to realistic and notional traffic and weather scenarios
– Effective in planning the management of departures & arrivals
– Evaluating complicating factors of uncertainty in operating conditions
32Final Briefing, 30 Sep 2016, Version 1
Conceptual
User Interface
33Final Briefing, 30 Sep 2016, Version 1
File View Window Help Current Time: 20452
RWY: 18L
Current Restrktions:
Model: ~
DMP·FIX V I .lml -.Emz
Model: ~
DMP · FIX V
Model: ~
DMP-FIX V
Highlight: PJ MERIL V I
Highlight: .
LILLS V
Highlight: i.
ANDYS, BUCKL, J... V
Avg Qlength
MaxQ length
Min Q length
UT: LT: TQL :
.E -• · 1 · 1 · 1 3 1 3 1 3 1 3 3 3 - -
2100 2200 2300
UT: LT: TQL :
.E -•••. 333333 ..
2100 2200 2300
UT: LT: TQL :
m m - - I - - 10 10 10 10 10 10 - •
Avg Off-Out (mins)
Max Off-Out (mins)
Min Off-Out (mins)
2100 2200 2300
B
Avg Off delay (mins) 38 Max Off delay (mins) 60 Min Off delay (mins) 30 25
<tJ-aoEING
~
Recommendations
What-if Operations
• Forecasting, what-if analysis and DMP implementation to
proactively minimize the negative impact of changing
weather, airport and traffic conditions
– Forecasting traffic flow restrictions, traffic conditions, and airport
operating conditions & estimating uncertainties of forecasts
– What-if analysis tool and process to design DMPs to
accommodate forecasts
– Categorical (fix specific) DMPs along with other runway-specific
DMPs or destination-specific TMIs for departures subject to
particular restrictions
– Collaboration of DRC with aircraft operators and other
stakeholders in the what-if analysis and DMP implementation
decision making
34Final Briefing, 30 Sep 2016, Version 1
Recommendations
What-if Tool
• Airport & airspace modeling
– Departure restrictions: other types, assignment to specific tail numbers
– Surface traffic interaction points which impede flow
– Gate modeling: assignment alternatives for arrivals, trail tracking for detailed
impact on aircraft utilization
– Variability in runway departure rates, transit times, gate occupancy times
– Verification: flight taxi times as per OOOI data, use delay fields from SWIM/FIXM
data as a source
• DMP emulation
– Individual runways, departure runways shared with arrivals
– Alternative implementation for distinct constraints
• Traffic & weather scenarios
– Additional scenarios including Lower visibility weather conditions at CLT, North
Flow runway operations, recovery from Ground Stop
– Design around traffic patterns of interest
35Final Briefing, 30 Sep 2016, Version 1
Backup
36Final Briefing, 30 Sep 2016, Version 1
Verification Results
Traffic Input To Simulation
37Final Briefing, 30 Sep 2016, Version 1
0 60 120 180 240 300 360 420 480 540 600 660 720 780 840 900 960 1020 1080
FAA ASPM 2 11 63 9 84 12 80 22 42 67 9 71 39 55 39 52 6 78 0
Simulation Input File 1 16 42 34 52 43 54 36 23 71 27 40 52 40 46 40 46 48 2
0
10
20
30
40
50
60
70
80
90
Nu
mb
er
of
De
pa
rtu
res
Time, Minutes
CLT Hourly Scheduled Departures, 8 Aug 2014
0 60 120 180 240 300 360 420 480 540 600 660 720 780 840 900 960 1020 1080
FAA ASPM 1 48 3 82 10 81 13 53 61 17 68 36 54 45 47 41 64 6 9
Simulation Input File 20 28 54 30 61 21 54 25 47 43 34 60 23 68 20 58 23 10 2
0
10
20
30
40
50
60
70
80
90
Nu
mb
er
of
Arr
iva
ls
Time, Minutes
CLT Scheduled Arrivals, 8 Aug 2014
• Differences between hourly counts of scheduled arrivals & departures
between FAA ASPM data and traffic schedule input to simulation– Traffic schedule input file derived from OOOI data, not schedule data
– May impact comparison of hourly statistics computed from simulation output data
<tJ-aoEING
I - • I -- I I I I I - • - I I I .L
I I l l T I I I I I I l l I l l l I 1 I I I I l T I I I I I l l
• I I I I I I I I I I I I I • I I I I I I I I I I I I • I I I I I I I I I I I I 7 • I I I I I I I I I I I I
Verification Results
Traffic Output From Simulation
38Final Briefing, 30 Sep 2016, Version 1
0 60 120 180 240 300 360 420 480 540 600 660 720 780 840 900 960 1020 1080
FAA ASPM 0 0 47 31 56 43 66 33 55 46 38 61 34 70 24 69 18 68 12
Simulation Output 6 23 34 36 45 42 50 45 32 54 38 36 48 46 43 42 53 37 3
0
10
20
30
40
50
60
70
80
Nu
mb
er
of
De
pa
rtu
res
Time, Minutes
CLT Departure Takeoffs, 8 Aug 2014
Source
Average Departure Throughput,
Departures Per Hour
Average Taxi Out Time,
Minutes
FAA ASPM, 8 August 2014 45 16.0
What-if Tool Traffic Simulation 36 15.9
Simulated vs. FAA ASPM hourly takeoff
rates differ significantly in many hours
Simulated vs. FAA ASPM hourly taxi-out
times comparable in many hours
• Simulated vs. FAA ASPM aggregate departure rates somewhat lower
• Simulated vs. FAA ASPM aggregate taxi-out times comparable
<tJ-aoEING
~
CLT Average Taxi Out Times, 8 Aug 2014
l fM ASPM • Smul.1tion Output
35.0 I
30.0
25.0
10.0
tS.o
10.0 -I L 5,0 - - f- -
I I I I I I I I I I I I I I I I I I I I• I I I I I I I I I I I I I I I I I I I I• I I I I I I I I I I I I I I I I I I I
0.0 0 I 60 I 110 I 1IIO I 140 I 300 j 360 410 I 4IIO 540 600
660 I :m 7IIO 840 h oo T 960 1020 11~ 1 r~ FAAASPM ll.O t 15.0 t 14.l t 16.1,t 16.0 14.0 15.5 14.l 16.1 16.0 12.1 17.7 16.1 19.l 14.4 t 16.9 1 ll.7 14.4 14.0
• smul.ation Output 9.9 11.4 15.3 11.6 11.61 11.6 r 18.8 17.1 14.5 I 18.5 15.4 1s.4 T ni 13.4 13.4 19.0 1 13.0 I 18.3 I 9..l
Traffic & Weather ScenariosBaseline Scenario Surface Traffic
39Final Briefing, 30 Sep 2016, Version 1
Scatter plot of actual departure taxi times by spot/runway assignment for
8/8/2014 baseline scenario
50 ~--~--------r-------,-----,----------,;::;-----,
45
'2 4o -E
~ 35 «I
ci3 30
>ro ~ 25 (I)
E F 20
~ I- 15
10
0 X
•
Spot25 Spot27 Spot29
0 •
O' 0 8 x is!
0
X 0 • 0
• 0
0
• 5 L------'------"------'-----'------'--~~~----
200 400 600 800 1000 1200 1400
Time of day (min )
<tJ-aoEING
~
70
I • Spot~ I X S pot9 X
60 X
-~ 50 -0 X X
6 a:) X .,... 40 X X
~ 1 X X ~ X •
x l1\ X9x • X X (I) 30 X X xJ ~ xX;;;:X ~ ~ E rf X
X
F X
»<X X fl ~
~ I x4 »< X ·x
] ' XIX ~~i XX* t ~ 20 xx
f_ X X n • ~ YX
t xx ~ 0 Xo X c(XX D X X f ~ X' X
10 • • x D x D • :«
• [] ox: X
0 L-----'---------'------"L---'-------'----'----'---~-~-~
400 500 600 700 600 900 1000 1100 1200 1300 1400
Time of day (min)
Traffic & Weather ScenariosBaseline Scenario Surface Traffic
40Final Briefing, 30 Sep 2016, Version 1
Scatter plot of actual departure taxi times overlaid onto histogram of gate
scheduled departure times for 8/8/2014 baseline scenario
Histogram: number of scheduled
departures by ¼ hour
Scatter Plot: departure taxi times for
potential future analysis & comparison
70 ~--~---~---~---~---~--~---~
~ 00 E dl E -F 50 -x m I- -+ 4.0 E m ._
~ 30 -u:i
..c u:i Clo .... -E LO
1-=i O 10
• . --• .. •
•
' ., • • ...
•
+
. • I
..
. II!
• • • .. •
0 '------LU.11-LLL.LLLJ..u..J.U..U-LLL.LLLJU..U.UlU-LJ..L.LLLJU..U..U...U-LJ..L.U..UU..U..LJ...J.JU.U.....J.LLJ..LJ..u..,_ __ __,
200 400 600 800 1000 1200 1400 1600
l i1me of day (min)
<tJ-aoEING
~
Verification Results
DMP Implementation
41Final Briefing, 30 Sep 2016, Version 1
• DMP results for simple test case hitl6-training-advisory.list_data– Throughput maintained
– Runway queue lengths comply with the target of 3/+1-2 aircraft
– Average quarter-hour taxi-out times departures are sharply reduced
– Taxi-out delay shifted to gate
Number of Actual Takeoffs 8
7
6
5
1 <l: 04 lii .0 E :, z
3
2
o ~~~~~~~~~
0 2 3 Simulated Time, Hours
25 Average Queue Length
20
<t: 15
~ <l: 0 lii .0 E :, Z 10
5
~
Simulated Time, Hours
70 Average Taxi Out Time
(/) Q)
60
50
"5 40 C:
~ Qi' E i= ·x 30 ~
20
10
~
o~~~~~~~~~
0 2 3
Simulated Time, Hours
<tJ-aoEING
~ '°/ifierage Gate Departure Delay
~ 50
40
20
10
o ~-~.J1~~~~~ 0 2 3
Simulated Time, Hours
Current User Interface
42Final Briefing, 30 Sep 2016, Version 1
Airport Demand Analysis DMP Analysis
<tJ-aoEING
~
~ Airport Dema •=+ l~ I liJ Departu l•= •I• 1...a.-1
Select initial scenario informat ion
Traffic Schedule File:!C:\Work\TerminalContro l\What If OY 1\Scen ariofiles\Sens itivity\hitl6-training- advisor Q Flow Restrictions File: l1inalContro l\Wh at If OY 1 \Scenariof iles\Sensitivity\Tra fficflowRestrictionsSh ort.csv IQ I Load Scenario ]
Choose parameters for DMP w hat-if analysis
Runw ay !Targ et Depa rture! Start Time Stop Time Queue Length
Simulation & Results Configuration 18C 3 70 183
Simulation Time Step (mins.)~ 18L 3 59 r s3
Plot Time-Bin Size (mins.) ~ Choose airport/airspace parameter values for demand w hat- if analysis I An alyze I
Runway
I Departu re Departure
I MIT I Start
I
End
Rate Fix Restri ction Time Time
18C 30 MERIL6 20 50 100
18L 30 BUCKL7 15 75 1135
[ A dd Restriction j [ Remove Restriction I I A nalyze I Demand analysis results
OMP w hat-if analysis results
Departure I Ave Throughput! Ave Queue Max Queue Min Queue IAve TaxiOut l Ave Gate Runway Length Length Len gth Tim e (min) Delay (min)
18C 29.2105 1.7418 6 0 16 .0840 7.4921
Departure I Ave Th roughput! Ave Queu e Max Qu eue I Min Qu eue IAve TaxiOut l Ave Gate Runway Length Length Length Tim e (min) Delay (min)
18L 27.8571 6.7857 15 0 31.3251 7.4979
18C 29.2105 2.9728 12 0 23.5761 8.6417e-16
18L 27.3684 9.0380 25 0 39.7462 1.0931e-15 A rrival I Ave Throughput! Ave Taxi In
Runw ay Time (min) 18R 29.6386 15.7073
23 26.5116 15.7368
Arrival I Ave Throughput! Ave Taxi In Runway Time (min)
18R 29.6386 12.0244
23 26.5116 12
I Configure OMP I I•
top related