A Simulation Study on a Method of Departure Taxi Scheduling at

Transcription

ENRI Int. Workshop on ATM/CNS (EIWAC2013) 19-21FEB2013
EN-024
A Simulation Study on a Method of
Departure Taxi Scheduling at
Haneda Airport
Izumi YAMADA,
Hisae AOYAMA, Mark BROWN,
Midori SUMIYA and Ryota MORI
ATM Department,ENRI
i-yamada
enri.go.jp
Outlines
Background and objectives
– Characteristics of Haneda airport
Traffic management algorithm
Results of fast-time simulation experiment
– Reduction of taxiing and queuing time
– Guarantee of takeoff time
Discussion
Conclusion
2
Backgrounds
Corresponding R&D vision
– ICAO ASBU(Aviation Systems Block Upgrades)
Module 80: Airport CDM (Collaborative Decision Making)
Module 15: AMAN/DMAN (Arrival/Departure Manager)
– Corresponding R&D reports in Europe and the United
States says…
Airport CDM is effective for improving efficiency and
punctuality of airport operation
– CARATS (JAPAN: Collaborative Actions for Renovation of Air Traffic
Systems)
Bottlenecks at congested airports and airspaces in the
Greater Tokyo Metropolitan area, etc. must be eliminated
3
Aims of the study
To examine a traffic management method
suitable for Haneda airport
– Departure taxi scheduling
Expected performance
1. Reduction of taxiing time
Especially for departure
2. Transparency in takeoff time planning and
execution (guarantee of takeoff time)
4
About Haneda airport (1/2)
The most congested airport in Japan
– Over 1,000 movements per day
– Origin and destination of major air traffic flow in
Japan
Sapporo
Mainly used for domestic airways
Osaka
Tokyo
(Haneda)
Fukuoka
Naha
5
About Haneda airport (2/2)
Int’l
terminal
Cargo
Domestic
terminal
A-RWY
3,000m
Hangar
Photo: MLITT
C-RWY
3,000m
Complex layout and operation
– 4 runways (2 pairs of parallel runways)
D-RWY
2,500m
3 or 4 runways constantly active
Interference between runways occurs frequently
– Gates: densely located around terminal buildings
6
Interference between runways
Example of southerly wind configuration
C-RWY
Gates
Domestic
terminal
H A N GA R
Crossing of
flight paths
A-RWY
International
terminal
N
Departure flight path
Arrival flight path
(considering go-around)
7
Simulated surface traffic flow
(simulator developed by ENRI)
Departure
Queuing Departure
Arrival
Runway occupation
Time(JST)
8
Location of congestion at Haneda
Almost limited in the area before departure
runway
– Relevant to apply taxi scheduling (queue
management) Departures (504 flights)
Mapping of taxiing time
with speed less than
10 [km/h]
(excl. pushback)
9
Simplified congestion model
Focusing on takeoff queues
– Dynamics of congestion will be determined by…
Takeoff capacity of runway system
Number of departures reaching takeoff queue
– Takeoff capacity drops temporarily
due to interference with arrival flow
Gates
Final
approach
Apron/
Taxiway
Takeoff queues
FIXED taxi times
w.r.t gate-runway pair
Runway system
10
Runway capacity constraint model
Based on Gilbo’s capacity model
– Count (#dep., #arr.) observations in 5 minute time
window, rejecting (0,0) as exception
– Evaluate the proportion of each (#dep., #arr.) in total
Arrivals
observations
on D-RWY
#dep.
#arr.
0
1
2
3
4
0
21.0%
11.6%
5.7%
1.2%
1
18.3% 17.1% 13.0%
5.4%
0
Departures
from A-RWY
2
3.9%
1.5%
1.0%
0
0
3
0
0
0
0
0
(Simulation example,
dep.: A-RWY, arr.: D-RWY)
Capacity constraint assumption
#
/5
3
#
/5
11
Departure scheduling algorithm 1/2
Taxi time
table
Predicted
flow
Runway capacity
model
Queue
prediction
Takeoff time
window
assignment
Block-out
time assignment
Final
approach
Arrivals
Gates
Apron/
Taxiway
Actual
flow
Takeoff queues
Departures
Runway system
12
Departure scheduling algorithm 2/2
a) Predicted interfering landing number per 5 minutes
One arrival within
5-minute window
… 18:45
Queue
prediction
3
2
… 18:45
19:30
… time
One departure within
13
5-minute window
9 12
19
8 11 15
18
22
5
21
4 6 7 10 14 16 17 20
19:00
19:15
19:30 … time
c) Takeoff time window assignment
3
2
1
Block-out time
assignment
19:15
b) Predicted takeoff demand per 5 minutes
1
Takeoff time
window
assignment
19:00
… 18:45
5
4
3
6
8 11 14 16 18 19
7 10 13 15 17
20 22
9 12
21
19:00
19:15
19:30 … time
13
Baseline scenario
Derived from observation of actual operation
– Block-out/ -in time and gate
– Takeoff / landing time and runway
Hourly traffic volume
80
arr22
Arr., B
arr23 D
Arr.,
Dep.,
dep16LC
Dep., A
number of flights
70
60
dep16R
50
40
Through the day
Dep.: 504 flights
Arr.: 525 flights
30
20
10
0
6
7
8
9
10
11
12
13
14
15
time (JST)
16
17
18
19
20
21
22
14
Modified scenario
94 departures were assigned block-out delay
Number of departures with
delayed block-out
14
7
for 16L: t_REQ moved
Delay-assigned
Dep., C-RWY
for 16R: t_REQ moved
Delay-assigned
Dep., A-RWY
Avg.average
of assigned
[min]
of assigneddelay
wait
12
10
6
5
8
4
6
3
4
2
2
1
0
0
6
7
8
9
10 11 12
13 14 15
time (JST)
16 17 18 19
Average of assigned delay [min]
– Sum of delay: 249 min.
– Many for congested period in the evening
20 21 22
15
Simulation result
Departures
continually
coming
Queue
Departure
Arrival
Reduced
queue
RWY16R
Baseline scenario
Modified scenario
Queue reduction in congested period (19:30 JST)
16
Performance index 1.
Reduction of taxiing time
How to measure
– Comparing taxiing/queuing time between the
simulation result of baseline and modified scenario
Baseline
Scenario
Taxiing time
DEP003
DEP002
DEP001
Queuing time
Modified
Scenario
DEP003
DEP002
DEP001
Block-out
Join queue
Takeoff
Time
17
Reduction of taxiing time
Significant reduction in the evening
– Through the day: total 2.12% (133 min.) reduction
for departure taxiing time
Decrease in departures' taxiing
time and queueing time [min]
55
50
decrease in taxiing time
45
decrease in queuing time
40
35
30
25
20
15
10
5
0
-5
-10
6
7
8
9
10
11
12
13
14
15
time (JST)
16
17
18
19
20
21
22
18
Guarantee of takeoff time
How to measure
– Punctuality: takeoff within the assigned takeoff
time window
Takeoff
5 minutes
Too early
Time
Assigned window
Takeoff
On time
Time
Assigned window
Takeoff
Too late
Time
Assigned window
19
Guarantee of takeoff time
63.3% of departures took off within assigned
time window
Too early
On time
Too late
90
80
Number of departures
70
60
92 acft.
(18.2%)
50
93 acft.
(18.5%)
319 acft.
(63.3%)
40
30
20
10
0
-5 -4 -3 -2 -1
0
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15+
Difference b/w planned and resulting takeoff time [min]
20
Discussion
18.2% took off slightly earlier than assigned
window (up to 4 min., mostly within 2 minutes)
– Due to rough assumption on runway capacity
Sometimes #dep. + #arr. > 3
18.5% took off later than assigned window
– In some cases, large deviation from assigned window
Though, takeoff times are same as baseline results
– Due to unmodeled congestion factor
Congestion at aprons
These may be solved by detailed modeling
21
Conclusions
Traffic management method suitable for
Haneda airport
– Departure taxi scheduling
Good performance obtained
– Reduction in departure taxi time : 2.12%
– Guarantee on takeoff time : 63.3%
Problems to be solved
– More precise forecast of runway capacity
– Taxi time prediction method considering
apron congestion
22
Acknowledgement
The authors express special thanks to
the Japanese Civil Aviation Bureau (JCAB)
of the Ministry of Land, Infrastructure,
Transport and Tourism (MLITT)
for providing the source data.
and especially…
Thank you for your attention!
i-yamada
enri.go.jp
23
24
BACKUP SLIDES
25
Motivation
CARATS says…
– “Bottlenecks at congested airports
and airspaces in the Greater
Tokyo Metropolitan area, etc.
must be eliminated”
Many literatures report
effectiveness of Airport CDM
– How will Airport CDM work at
Haneda airport?
Reference: JCAB, “Long-term Vision for the Future Air Traffic Systems”, 2010.
http://www.mlit.go.jp/common/000128185.pdf
26
Our research topics
Technical arguments on traffic management at
Haneda airport
– Post-operation data processing
– Surface traffic flow analysis
Identification of congested area
Queue analysis
– Airport surface movement simulator
– Traffic management methods
– Evaluation methods for traffic management
27
Traffic management algorithm 1/2
Arrivals assumed as independent movement
– Landing time assumed as fixed
enabling takeoff capacity prospect
Time management for departures
– Predict takeoff demand at runway from initial
planning of departing gate
– Detect excess demand compared to the prospect of
takeoff capacity
– Assign wait at gate for excess demand
28
Congestion at apron
Departure’s taxiing route is blocked by arrivals
headway
blocked
29

A Simulation Study on a Method of Departure Taxi Scheduling at

Transcription

Similar documents

embraer authorized service center