A Parametric Analysis of Air Traffic Flow Control Options Under Weather Uncertainty November 17, 2002 INFORMS Annual Meeting Lisa Grignon, IE, University of Washington Aslaug Haraldsdottir, ATM Group, The Boeing Company Joyce Yen, IE, University of Washington Zelda Zabinsky, IE, University of Washington
28
Embed
A Parametric Analysis of Air Traffic Flow Control Options Under Weather Uncertainty
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
A Parametric Analysis of Air Traffic Flow Control Options Under Weather Uncertainty
November 17, 2002
INFORMS Annual Meeting
Lisa Grignon, IE, University of Washington
Aslaug Haraldsdottir, ATM Group, The Boeing Company
Joyce Yen, IE, University of Washington
Zelda Zabinsky, IE, University of Washington
November 17, 2002 INFORMS Conference 2
Objective
Conduct a parametric analysis on flow control options to gain a better understanding of
their effects on a dynamic air traffic system
We would like to compare the trade-off between ground delay and air delay given
uncertainties in the weather prediction
November 17, 2002 INFORMS Conference 3
Agenda
• Air Traffic Background
• Simple Policy-Based Approach
• Stochastic Optimization Formulation
• Next Steps
November 17, 2002 INFORMS Conference 4
Flow Control Decisions
• A collaborative decision is made between Air Traffic Control (ATC), the Airline Operational Control (AOC), and affected centers
• Flow control options result in either some form of ground delay or air delay
• Two major flow control options– Ground holding (delay on the ground)
– Miles-in-Trail (delay in the air)
November 17, 2002 INFORMS Conference 5
Question
How do we make decisions regarding ground delay and air delay to minimize
either the total delay or the cost of delay?
November 17, 2002 INFORMS Conference 6
Probabilistic Problem• Consider a set of planes arriving at a single airport
N
N-1
2
1
AirportQueue
(RunwaySystem)
Flights
Queuing Phenomenon
Due to weather uncertainty, there is a probabilistic reduction of capacity, airport arrival rate (AAR)
November 17, 2002 INFORMS Conference 7
Simple Policy-Based Approach to Determine Ground and Air Delay
• Reschedule the flights based on the probabilistic arrival rate and calculate the ground delay
• Using the arrival rate profiles, assess the actual landing times and calculate the expected air delay, using FIFO policy
• Use these calculations to estimate the expected total delay
November 17, 2002 INFORMS Conference 8
Probabilistic Arrival Rate Example
Probability Time 1 Time 2 Time 3 Time 4
Demand 30 40 20 0
AAR Profile 1 0.5 10 30 40 90
AAR Profile 2 0.1 30 10 40 90
AAR Profile 3 0.4 10 10 40 90
Original AAR 40 40 40 90
Most Likely AAR 10 30 40 90
Expected AAR 12 20 40 90
November 17, 2002 INFORMS Conference 9
Expected Delay Under Three Policy Decisions
0
10
20
30
40
50
60
70
80
90
Original AAR Most Likely AAR Expected AAR
Del
ay E[Air Delay]
Ground Delay
November 17, 2002 INFORMS Conference 10
Cost of Delay
0.00
50.00
100.00
150.00
200.00
250.00
300.00
350.00
400.00
Air $ = Ground $ Air $ = 2xGround $ Air $ = 5xGround $
Co
st Original AAR
Most Likely AAR
Expected AAR
November 17, 2002 INFORMS Conference 11
Summary of Policy-Based Approach
A more conservative decision (lower AAR) results in a higher expected delay, but may result in lower expected costs
Although we can choose the best of the three policies based on cost information, not all possible policies have been considered
November 17, 2002 INFORMS Conference 12
Stochastic Optimization Formulation
• Pose as a stochastic optimization problem
• Consider flights arriving at a single airport
• Aggregate flights in groups based upon the original scheduled arrival
• Octavio Richetta and Amedeo Odoni (1993,1994)– Min E[Cost of ground delay] + E[Cost of air delay]
– Dynamic formulation
November 17, 2002 INFORMS Conference 13
Modification of Objective Function
• In addition to cost of ground delay and air delay, the value of the system should include the utility of the flights based on their total delay
• This new objective would be a utilitarian point of view; good for both ATC and AOC
November 17, 2002 INFORMS Conference 14
Utility of Total Delay
• Total delay is important
• Utility of a flight is based on total delay
• A two hour delay results in 50% utility
Delay Time of Flight
Utility ofFlight
100 %
50 %
2 hours
Utility of a Flight as a function of Total Delay
November 17, 2002 INFORMS Conference 15
Modified Stochastic Optimization Problem
Decision variables determine the AAR for each time period and distribute the arrival rate to flights according to their originally scheduled arrival time
First stage decisions (Xij) reschedule the arrival time of flights from i to j
Recourse decisions (Ωijk) assign actual arrival time k (which may differ from the original arrival time i or rescheduled arrival time j)
November 17, 2002 INFORMS Conference 16
Mathematical Formulation (TEX slide)
kjiSX
TkX
TjiX
TkMS
MS
TiNX
ijCjkCikU
ijkkij
k
i
k
ijijkij
T
jkijkij
k
ik
k
ijijkk
T
ii
T
ii
i
T
ijij
T
i
T
ij
T
jkgaijk
,,0,,
1,,10)(
,,1,
1,,1
,,1
subject to
)]()()([max
1
1
1
11
1
1
1 1
November 17, 2002 INFORMS Conference 17
Example Problem
Probability Time 1 Time 2 Time 3 Time 4 *Demand 30 40 20 0