Surface Analysis to Support AEDT Aircraft Performance ......performance modeling in AEDT in order to...
Transcript of Surface Analysis to Support AEDT Aircraft Performance ......performance modeling in AEDT in order to...
![Page 1: Surface Analysis to Support AEDT Aircraft Performance ......performance modeling in AEDT in order to better reflect actual operations • Develop and validate enhanced taxi models](https://reader030.fdocuments.in/reader030/viewer/2022041102/5f01cb1e7e708231d4011073/html5/thumbnails/1.jpg)
FAA CENTER OF EXCELLENCE FOR ALTERNATIVE JET FUELS & ENVIRONMENT
Project manager: Hua He and Aniel Jardines, FAA Lead investigators: Hamsa Balakrishnan (MIT) and Tom Reynolds (MIT LL)
Surface Analysis to Support AEDT Aircraft Performance Module Development
Project 46
April 18-19, 2017 Alexandria, VA
Opinions, findings, conclusions and recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of ASCENT sponsor organizations.
![Page 2: Surface Analysis to Support AEDT Aircraft Performance ......performance modeling in AEDT in order to better reflect actual operations • Develop and validate enhanced taxi models](https://reader030.fdocuments.in/reader030/viewer/2022041102/5f01cb1e7e708231d4011073/html5/thumbnails/2.jpg)
2
Motivation
• Taxi phase in the Aviation Environmental Design Tool (AEDT) is currently modeled using default or user-specified taxi times, coupled with engine idle fuel and emissions assumptions from ICAO Aircraft Engine Emissions Databank
• These assumptions reduce the accuracy of the taxi performance modeling
• For some applications, higher fidelity surface modeling approaches may be needed
![Page 3: Surface Analysis to Support AEDT Aircraft Performance ......performance modeling in AEDT in order to better reflect actual operations • Develop and validate enhanced taxi models](https://reader030.fdocuments.in/reader030/viewer/2022041102/5f01cb1e7e708231d4011073/html5/thumbnails/3.jpg)
3
Objectives
• Identify needs and evaluate methods for improving taxi performance modeling in AEDT in order to better reflect actual operations
• Develop and validate enhanced taxi models by combining ASDE-X surface track data with statistical models of engine performance (using Flight Data Recorder information)
• Make recommendations for enhancements to AEDT Aircraft Performance Module (APM) based on knowledge gained
![Page 4: Surface Analysis to Support AEDT Aircraft Performance ......performance modeling in AEDT in order to better reflect actual operations • Develop and validate enhanced taxi models](https://reader030.fdocuments.in/reader030/viewer/2022041102/5f01cb1e7e708231d4011073/html5/thumbnails/4.jpg)
4
Outcomes and Practical Applications
AEDT APM
Total fuel burn
per time-period, per airport
Fuel flow rate profiles
ASDE-X data
Current scope
![Page 5: Surface Analysis to Support AEDT Aircraft Performance ......performance modeling in AEDT in order to better reflect actual operations • Develop and validate enhanced taxi models](https://reader030.fdocuments.in/reader030/viewer/2022041102/5f01cb1e7e708231d4011073/html5/thumbnails/5.jpg)
5
Outcomes and Practical Applications
AEDT APM
Total fuel burn, noise and emissions
per time-period, per airport
Fuel flow rate profiles
Thrust profiles
ASDE-X data
Potential extensions
![Page 6: Surface Analysis to Support AEDT Aircraft Performance ......performance modeling in AEDT in order to better reflect actual operations • Develop and validate enhanced taxi models](https://reader030.fdocuments.in/reader030/viewer/2022041102/5f01cb1e7e708231d4011073/html5/thumbnails/6.jpg)
6
Approach
AEDT APM
surface model needs
assessment
Aircraft surface
performance modeling
enhancements
Aircraft taxi performance
model validation
AEDT APM
enhancement recommendations
Stakeholder input,
supporting documents & prior research
ASDE-X data
FDR data
Model development data Validation data
![Page 7: Surface Analysis to Support AEDT Aircraft Performance ......performance modeling in AEDT in order to better reflect actual operations • Develop and validate enhanced taxi models](https://reader030.fdocuments.in/reader030/viewer/2022041102/5f01cb1e7e708231d4011073/html5/thumbnails/7.jpg)
7
Schedule and Status
• Assess AEDT aircraft surface performance modeling needs
• Develop enhanced aircraft surface performance models
• Validate enhanced aircraft surface performance models
• Recommend AEDT APM enhancements
[Sept.-Nov. 2016]
[Oct. 2016-present]
[Jan. 2017-present]
[~Aug. 2017]
![Page 8: Surface Analysis to Support AEDT Aircraft Performance ......performance modeling in AEDT in order to better reflect actual operations • Develop and validate enhanced taxi models](https://reader030.fdocuments.in/reader030/viewer/2022041102/5f01cb1e7e708231d4011073/html5/thumbnails/8.jpg)
8
Recent Accomplishments [1]
• Assessment of AEDT surface APM modeling needs – Synthesis of findings from AEDT documentation, ACRP studies
02-45 and 02-27, and stakeholder input • Airport average taxi-time is either default (7 min taxi-in/ 19 min taxi-out)
or airport-specific for 75 airports – Default taxi database is outdated – Analysis in ACRP 02-45 suggests 16 min for taxi-out/ 7 min for taxi-in – Airport specific taxi-time estimates off by 4-50% [recent ASPM data]
• Taxiway network model of airport; taxi-path for each flight – Taxi-time based on taxi-path length and speed (AEDT default: 15 knots)
» ACRP report suggests ~13 knots – Taxi-in time is considered to be unimpeded; taxi-out time includes a queuing
delay at the runway. Queuing delay is sum of runway occupancy times of aircraft already in the queue
» Queues are assumed only at the runway; effects of acceleration not included
» Deterministic runway occupancy times. Variation by aircraft type/weather condition not known
• No surface-specific regression model for fuel flow rates
![Page 9: Surface Analysis to Support AEDT Aircraft Performance ......performance modeling in AEDT in order to better reflect actual operations • Develop and validate enhanced taxi models](https://reader030.fdocuments.in/reader030/viewer/2022041102/5f01cb1e7e708231d4011073/html5/thumbnails/9.jpg)
9
Recent Accomplishments [2]
• Assessment of AEDT surface APM modeling needs
• Identified needs: – Models that are representative of a wide range of taxi
conditions, aircraft types, airports, airlines, and weather conditions
– Identify key locations and events at major airports (e.g., non-movement area, taxiway intersections, spot and departure runway queues, runway crossings, etc.)
– Need to model operational variability • Assess uncertainty associated with fuel burn estimates • Evaluate sensitivity to various factors (e.g., takeoff mass,
ambient conditions) – Need data-driven validation of models
![Page 10: Surface Analysis to Support AEDT Aircraft Performance ......performance modeling in AEDT in order to better reflect actual operations • Develop and validate enhanced taxi models](https://reader030.fdocuments.in/reader030/viewer/2022041102/5f01cb1e7e708231d4011073/html5/thumbnails/10.jpg)
10
Recent Accomplishments [3]
• Surface APM model enhancements – In order to determine the “full” fuel burn
profile, need to consider both the pre-ASDE-X (frequently the non-movement area) portion and the ASDE-X portion of surface operations
– The pre-ASDE-X portion may include engine start and ramp area movements
– Can vary by airport; needs characterization by airport
– For the ASDE-X portion, prior work suggests that the taxi time and number of acceleration events are significant
Extract taxi time & # of
accelerations Fuel burn estimate
Regression model
ASDE-X
Gate
Enginestart
Taxi
Push-back
FDR dataASDE-X data
Pre-ASDE-X contribution
![Page 11: Surface Analysis to Support AEDT Aircraft Performance ......performance modeling in AEDT in order to better reflect actual operations • Develop and validate enhanced taxi models](https://reader030.fdocuments.in/reader030/viewer/2022041102/5f01cb1e7e708231d4011073/html5/thumbnails/11.jpg)
11
Recent Accomplishments [4]
• Initial identification of Non-Movement Area (NMA) & “pre-ASDE-X” area (PAXA) for US airports in FDR data sets of European carrier (BOS, JFK, EWR, MIA, ORD, LAX) – Determine NMA (based on airport markings) & PAXA (based on
ASDE-X data analysis) polygons for each of the US airports
BOS JFK
MIA ORD
NMAPAXA
![Page 12: Surface Analysis to Support AEDT Aircraft Performance ......performance modeling in AEDT in order to better reflect actual operations • Develop and validate enhanced taxi models](https://reader030.fdocuments.in/reader030/viewer/2022041102/5f01cb1e7e708231d4011073/html5/thumbnails/12.jpg)
12
Recent Accomplishments [5]
• Determining PAXA regions
ASDE-X Archives
Select airport, analysis days & plot
data
Generate coverage grid
Identify PAXA polygon based on interior edges of largest cluster
Latit
ude
Longitude
Latit
ude
Longitude
e.g. MIA
e.g., MIA
![Page 13: Surface Analysis to Support AEDT Aircraft Performance ......performance modeling in AEDT in order to better reflect actual operations • Develop and validate enhanced taxi models](https://reader030.fdocuments.in/reader030/viewer/2022041102/5f01cb1e7e708231d4011073/html5/thumbnails/13.jpg)
13 750 800 850 900 950Time (s)
-2
0
2
4
6
8
10
Spee
d (m
/s)
DataFilterSmoother
Recent Accomplishments [6]
• Smoothing/filtering algorithms for FDR and ASDE-X data – Need to “match” events (e.g., turns and accelerations) – Resolution and noise characteristics vary between datasets – Infer ground truth data (equivalent tracks) for algorithm validation
100 200 300 400 500 600 700Time
0
2
4
6
8
10
12
Spee
d (m
/s)
DataFilterSmootherASDE-XFDR
(Not the same flight)
![Page 14: Surface Analysis to Support AEDT Aircraft Performance ......performance modeling in AEDT in order to better reflect actual operations • Develop and validate enhanced taxi models](https://reader030.fdocuments.in/reader030/viewer/2022041102/5f01cb1e7e708231d4011073/html5/thumbnails/14.jpg)
14
Summary
• Development and validation of enhancements to AEDT surface APM
• Next steps – Model range of taxi conditions, aircraft types, airports, airlines, and
weather conditions – Analyze ASDE-X data from major airports to identify key locations
and events (e.g., non-movement area, taxiway intersections, spot and departure runway queues, runway crossings, etc.)
– Model operational variability and evaluate sensitivity – Develop and validate models by synthesizing data from
• 2006 FDR (European carrier) • 2016 FDR (European carrier, including 123 arrivals and 368 departures
at six US airports) • Aggregate (i.e., total over taxi phase) fuel use data from A4A • 2016 ASDE-X data
![Page 15: Surface Analysis to Support AEDT Aircraft Performance ......performance modeling in AEDT in order to better reflect actual operations • Develop and validate enhanced taxi models](https://reader030.fdocuments.in/reader030/viewer/2022041102/5f01cb1e7e708231d4011073/html5/thumbnails/15.jpg)
15
References
• ACRP 02-45, “Methodology to Improve EDMS/AEDT Quantification of Aircraft Taxi/Idle Emissions”, Transportation Research Board, 2016.
• ACRP 02-27, “Aircraft Taxi Noise Database for Airport Noise Modeling”, Transp. Research Board, 2013.
• Chati, Y.S., and Balakrishnan, H., “Analysis of Aircraft Fuel Burn and Emissions in the Landing and Takeoff Cycle using Operational Data," Intl. Conf. on Research in Air Transportation, Istanbul, 2014.
• DOT-VNTSC-FAA-16-11, “Aviation Environmental Design Tool – Technical manual, Version 2b”, U.S Department of Transportation (FAA), 2016.
• H. Khadilkar and H. Balakrishnan, “Estimation of Aircraft Taxi Fuel Consumption using Flight Data Recorder Archives”, Transportation Research Part D: Transport and the Environment, 17, (7), 2012.
• Särkkä, Simo. "Unscented Rauch-Tung-Striebel Smoother." IEEE Transactions on Automatic Control 53.3 (2008): 845-849.
• Lopez, Remy, and Patrick Danes. "Low-Complexity IMM Smoothing for Jump Markov Nonlinear Systems." IEEE Transactions on Aerospace and Electronic Systems (2017).
Contributors • PIs: Hamsa Balakrishnan and Tom Reynolds • MIT Lincoln Laboratory: Emily Clemons • MIT Students: Sandeep Badrinath, Yashovardhan Chati • FAA: Chris Dorbian
![Page 16: Surface Analysis to Support AEDT Aircraft Performance ......performance modeling in AEDT in order to better reflect actual operations • Develop and validate enhanced taxi models](https://reader030.fdocuments.in/reader030/viewer/2022041102/5f01cb1e7e708231d4011073/html5/thumbnails/16.jpg)
16