Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management
-
Upload
mbloem -
Category
Engineering
-
view
87 -
download
4
Transcript of Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management
![Page 1: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/1.jpg)
Optimization and Analytics
for Air Traffic Management
Enabling Decision-Support Tools
Michael Bloem
Stanford University NASA Ames Research Center
![Page 2: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/2.jpg)
Outline
• Air traffic management background
• Research topics
• Research objective and approach
• Case study:
decision-support tool for area supervisors
• Summary of contributions
2
![Page 3: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/3.jpg)
Air Traffic Management is Important
3
![Page 4: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/4.jpg)
Air Traffic Management is Important
civil aviation
was responsible for
5.4% of US GDP
in 2012
[FAA’s 2014 "The Economic Impact of Civil Aviation on the US Economy"]
3
![Page 5: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/5.jpg)
Air Traffic Management is Important
civil aviation
was responsible for
5.4% of US GDP
in 2012
837 million passengers
carried by airlines
operating in US airspace
in 2012
[FAA’s 2014 "The Economic Impact of Civil Aviation on the US Economy"]
3
![Page 6: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/6.jpg)
Air Traffic Management is Complex
4
![Page 7: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/7.jpg)
Air Traffic Management is Complex
∼50,000 flights/day
5,000+ flights simultaneously
4
![Page 8: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/8.jpg)
Air Traffic Management is
Accomplished Largely by
Human Decision Makers
5
![Page 9: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/9.jpg)
Air Traffic Management is
Accomplished Largely by
Human Decision Makers
5
![Page 10: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/10.jpg)
Air Traffic Management is
Accomplished Largely by
Human Decision Makers
5
![Page 11: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/11.jpg)
Air Traffic Management is
Accomplished Largely by
Human Decision Makers
5
![Page 12: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/12.jpg)
Air Traffic Management is
Accomplished Largely by
Human Decision Makers
5
![Page 13: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/13.jpg)
Research Topics: ATM Decisions
6
![Page 14: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/14.jpg)
Research Topics: ATM Decisions
1 Area supervisors:area configuration selection
– Motivation: prescriptive decision model
6
![Page 15: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/15.jpg)
Research Topics: ATM Decisions
1 Area supervisors:area configuration selection
– Motivation: prescriptive decision model
2 Operations managers:assignment of flights to slots
– Motivation: insight into airline delay costs
6
![Page 16: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/16.jpg)
Research Topics: ATM Decisions
1 Area supervisors:area configuration selection
– Motivation: prescriptive decision model
2 Operations managers:assignment of flights to slots
– Motivation: insight into airline delay costs
3 Flow managers:Ground Delay Program implementation
– Motivation: predictive capability and insights
6
![Page 17: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/17.jpg)
Research Objective
decision-
support
tool
7
![Page 18: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/18.jpg)
Research Objective
solution
algorithm
decision-
support
tool
decision
model
constraints
objective
7
![Page 19: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/19.jpg)
Research Approach
expert input
& feedback
solution
algorithm
decision-
support
tool
decision
model
constraints
objective
8
![Page 20: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/20.jpg)
Research Approach
expert input
& feedback
operational
decision
data
solution
algorithm
decision-
support
tool
decision
model
constraints
objective
8
![Page 21: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/21.jpg)
Research Approach
fast-time
simulations
expert input
& feedback
operational
decision
data
solution
algorithm
decision-
support
tool
decision
model
constraints
objective
8
![Page 22: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/22.jpg)
Research Approach
fast-time
simulations
human-in-
the-loop
simulations
expert input
& feedback
operational
decision
data
solution
algorithm
decision-
support
tool
decision
model
constraints
objective
8
![Page 23: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/23.jpg)
Outline
• Air traffic management background
• Research topics
• Research objective and approach
• Case study:
decision-support tool for area supervisors
• Summary of contributions
9
![Page 24: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/24.jpg)
En-Route Air Traffic Control Centers
10
![Page 25: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/25.jpg)
En-Route Air Traffic Control Centers
10
![Page 26: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/26.jpg)
Cleveland Center
11
![Page 27: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/27.jpg)
Cleveland Center Area 4
4645
12
![Page 28: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/28.jpg)
Cleveland Center Area 4
47
48
49
4546
36,000
31,000
24,000Longitude
Latitude
Altitude
13
![Page 29: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/29.jpg)
Sample Area of Specialization Resources
Sectors
47
48
49
4546 4546
4749
48
14
![Page 30: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/30.jpg)
Sample Area of Specialization Resources
Sectors
Controllers
available to fill
operating
positions
47
48
49
4546 4546
4749
48
09:00–09:30 :
09:30–11:00 :
14
![Page 31: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/31.jpg)
Sample Area of Specialization Resources
Sectors
Controllers
available to fill
operating
positions
Workstations
47
48
49
4546 4546
4749
48
09:00–09:30 :
09:30–11:00 :
48
14
![Page 32: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/32.jpg)
Modeling Decisions and Developing Algorithms
fast-time
simulations
human-in-
the-loop
simulations
expert input
& feedback
operational
decision
data
solution
algorithm
decision-
support
tool
decision
model
constraints
objective
15
![Page 33: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/33.jpg)
Modeling Decisions and Developing Algorithms
fast-time
simulations
15
![Page 34: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/34.jpg)
Sample Sectors, Configuration, and Traffic
sample sectors:
16
![Page 35: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/35.jpg)
Sample Sectors, Configuration, and Traffic
sample sectors: at time step k:
Ck
16
![Page 36: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/36.jpg)
Sample Sectors, Configuration, and Traffic
Tk
sample sectors: at time step k:
Ck
16
![Page 37: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/37.jpg)
Decision Model
Configuration Schedule Advisory Problem (CSA)
minimize
K∑
k=1
gk(Ck−1, Tk−1, Ck, Tk)
subject to Ck ∈ Ck, k = 0,1,2, . . . , K
17
![Page 38: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/38.jpg)
Decision Model
Configuration Schedule Advisory Problem (CSA)
minimize
K∑
k=1
gk(Ck−1, Tk−1, Ck, Tk)
subject to Ck ∈ Ck, k = 0,1,2, . . . , K
gk: single-time step advisory cost
17
![Page 39: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/39.jpg)
Decision Model
Configuration Schedule Advisory Problem (CSA)
minimize
K∑
k=1
gk(Ck−1, Tk−1, Ck, Tk)
subject to Ck ∈ Ck, k = 0,1,2, . . . , K
gk: single-time step advisory cost
Ck: set of valid configurations at k
17
![Page 40: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/40.jpg)
Sample Scenario
0timestep1 2 3 4 5
18
![Page 41: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/41.jpg)
Sample Scenario
0timestep1 2 3 4 5
18
![Page 42: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/42.jpg)
Sample Scenario
0timestep1 2 3 4 5
18
![Page 43: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/43.jpg)
Sample Scenario
0timestep1 2 3 4 5
18
![Page 44: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/44.jpg)
Sample Scenario
0timestep1 2 3 4 5
18
![Page 45: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/45.jpg)
Sample Scenario
0timestep1 2 3 4 5
18
![Page 46: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/46.jpg)
Configuration Constraints
0timestep1 2 3 4 5
19
![Page 47: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/47.jpg)
Configuration Constraints
0timestep1 2 3 4 5
19
![Page 48: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/48.jpg)
Configuration Constraints
0timestep1 2 3 4 5
C2
19
![Page 49: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/49.jpg)
Possible Algorithm Advisory
0timestep1 2 3 4 5
20
![Page 50: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/50.jpg)
Possible Algorithm Advisories
0timestep1 2 3 4 5
21
![Page 51: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/51.jpg)
Possible Algorithm Advisories
0timestep1 2 3 4 5
25 = 32 possibilities
21
![Page 52: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/52.jpg)
Advisory Cost: Static Cost
gSk(Ck, Tk)
20% 40% 60% 80% 100% 120%0
2
4
6
8
10
Open Sector Load[aircraft count/Monitor Alert Parameter]
Two−operating positionstatic costgS
k(Ck, Tk)
22
![Page 53: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/53.jpg)
Advisory Cost: Static Cost
gSk(Ck, Tk)
20% 40% 60% 80% 100% 120%0
2
4
6
8
10
Open Sector Load[aircraft count/Monitor Alert Parameter]
Two−operating positionstatic cost
One−operating positionstatic cost
gSk(Ck, Tk)
22
![Page 54: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/54.jpg)
Advisory Cost: Reconfiguration Cost
gRk(Ck−1, Tk−1,Ck, Tk)
(1)
(1)
(2)
(1)
(2)
(2)
(1)
Ck−1, Tk−1 Ck, Tk
23
![Page 55: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/55.jpg)
Advisory Cost: Reconfiguration Cost
gRk(Ck−1, Tk−1,Ck, Tk)
sector & flights moving
(1)
(1)
(2)
(1)
(2)
(2)
(1)
Ck−1, Tk−1 Ck, Tk
between workstations
23
![Page 56: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/56.jpg)
Advisory Cost: Reconfiguration Cost
gRk(Ck−1, Tk−1,Ck, Tk)
open sector gaining
sector & flights moving
(1)
(1)
(2)
(1)
(2)
(2)
(1)
Ck−1, Tk−1 Ck, Tk
between workstations
2nd operating position
23
![Page 57: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/57.jpg)
Advisory Cost: Reconfiguration Cost
gRk(Ck−1, Tk−1,Ck, Tk)
open sector gaining
open sector losingsector & flights moving
(1)
(1)
(2)
(1)
(2)
(2)
(1)
Ck−1, Tk−1 Ck, Tk
between workstations 2nd operating position
2nd operating position
23
![Page 58: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/58.jpg)
Advisory Cost: Reconfiguration Cost
gRk(Ck−1, Tk−1,Ck, Tk)
open sector gaining
open sector losingsector & flights moving
(1)
(1)
(2)
(1)
(2)
(2)
(1)
Ck−1, Tk−1 Ck, Tk
between workstations 2nd operating position
2nd operating position
single-time step advisory cost:
gk(Ck−1, Tk−1, Ck, Tk) = gSk(Ck, Tk)+β
RgRk(Ck−1, Tk−1, Ck, Tk)
23
![Page 59: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/59.jpg)
Advisory Cost: Reconfiguration Cost
gRk(Ck−1, Tk−1,Ck, Tk)
open sector gaining
open sector losingsector & flights moving
(1)
(1)
(2)
(1)
(2)
(2)
(1)
Ck−1, Tk−1 Ck, Tk
between workstations 2nd operating position
2nd operating position
single-time step advisory cost:
gk(Ck−1, Tk−1, Ck, Tk) = gSk(Ck, Tk)+β
RgRk(Ck−1, Tk−1, Ck, Tk)
operational decision data & fast-time simulations
leveraged to select βR
23
![Page 60: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/60.jpg)
Minimum-Cost Advisory
0timestep1 2 3 4 5
0
00 3
0 1
0 0
8
5
0
24
![Page 61: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/61.jpg)
Minimum-Cost Advisory
0timestep1 2 3 4 5
0
00 3
0 1
0 0
8
5
0
gSk(C1, T1)
24
![Page 62: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/62.jpg)
Minimum-Cost Advisory
0timestep1 2 3 4 5
0
00 3
0 1
0 0
8
5
0
gRk(C3, T3, C4, T4)
gSk(C1, T1)
24
![Page 63: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/63.jpg)
Decision Model and Solution Algorithm
Configuration Schedule Advisory Problem (CSA)
minimize
K∑
k=1
gSk(Ck, Tk) + βRgR
k(Ck−1, Tk−1, Ck, Tk)
subject to Ck ∈ Ck, k = 0,1,2, . . . , K
25
![Page 64: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/64.jpg)
Decision Model and Solution Algorithm
Configuration Schedule Advisory Problem (CSA)
minimize g(C,T)
subject to Ck ∈ Ck, k = 0,1,2, . . . , K
25
![Page 65: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/65.jpg)
Decision Model and Solution Algorithm
Configuration Schedule Advisory Problem (CSA)
minimize g(C,T)
subject to Ck ∈ Ck, k = 0,1,2, . . . , K
Shortest path problem → use A∗ algorithm
25
![Page 66: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/66.jpg)
Decision Model Relative to Previous Research:
Objectives and Constraints
AdvisoryCharacteristic
matchconfigurations to
traffic
few disruptivereconfigurations
number of opensectors
number ofoperatingpositions
26
![Page 67: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/67.jpg)
Decision Model Relative to Previous Research:
Objectives and Constraints
AdvisoryCharacteristic
Cano et al.(2007)
MB et al.(2008–2009)
Tien et al.(2010–2012)
matchconfigurations to
traffic
constraintand
optimizeconstraint constraint
few disruptivereconfigurations
constraintnot
consideredconstraint
number of opensectors
minimize minimizenot
considered
number ofoperatingpositions
notmodeled
not modeled minimize
26
![Page 68: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/68.jpg)
Decision Model Relative to Previous Research:
Objectives and Constraints
AdvisoryCharacteristic
Cano et al.(2007)
MB et al.(2008–2009)
Tien et al.(2010–2012)
matchconfigurations to
traffic
constraintand
optimizeconstraint constraint
few disruptivereconfigurations
constraintnot
consideredconstraint
number of opensectors
minimize minimizenot
considered
number ofoperatingpositions
notmodeled
not modeled minimize
26
![Page 69: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/69.jpg)
Decision Model Relative to Previous Research:
Objectives and Constraints
AdvisoryCharacteristic
Cano et al.(2007)
MB et al.(2008–2009)
Tien et al.(2010–2012)
DecisionModel
matchconfigurations to
traffic
constraintand
optimizeconstraint constraint optimize
few disruptivereconfigurations
constraintnot
consideredconstraint optimize
number of opensectors
minimize minimizenot
consideredconstraint
number ofoperatingpositions
notmodeled
not modeled minimize constraint
26
![Page 70: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/70.jpg)
Decision Model Relative to Previous Research:
Objectives and Constraints
AdvisoryCharacteristic
Cano et al.(2007)
MB et al.(2008–2009)
Tien et al.(2010–2012)
DecisionModel
matchconfigurations to
traffic
constraintand
optimizeconstraint constraint optimize
few disruptivereconfigurations
constraintnot
consideredconstraint optimize
number of opensectors
minimize minimizenot
consideredconstraint
number ofoperatingpositions
notmodeled
not modeled minimize constraint
26
![Page 71: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/71.jpg)
Decision Model vs. Operational Decisions:
Problem Instances for Fast-Time Simulations
47
48
49
4546 4546
4749
48
• traffic and configurations from
231 days in 2011 and 2012
• 6 am to midnight local time
• rolling horizon: implement first
hour of two-hour advisories
• five-minute time steps
27
![Page 72: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/72.jpg)
Decision Model vs. Operational Decisions:
Few Disruptive Reconfigurations
3 6 9 12 150
500
1000
1500
2000
2500
Number ofOpen Sector
Instances
Duration [hours]
operational
model
28
![Page 73: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/73.jpg)
Decision Model vs. Operational Decisions:
Match Configurations to Traffic
20% 40% 60% 80% 100% 120%0
2
4
6
8
10
Open Sector Load[aircraft count/Monitor Alert Parameter]
Low HighJust Right
gSk(Ck, Tk)
29
![Page 74: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/74.jpg)
Decision Model vs. Operational Decisions:
Match Configurations to Traffic
Low Just Right High0
20
40
60
80
100
Percent ofOpen Sector-
Minutes
operational
model
30
![Page 75: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/75.jpg)
Modeling Decisions and Developing Algorithms
fast-time
simulations
31
![Page 76: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/76.jpg)
Modeling Decisions and Developing Algorithms
fast-time
simulations
31
![Page 77: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/77.jpg)
Experts’ Feedback and Suggestion
Limitations of decision model
32
![Page 78: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/78.jpg)
Experts’ Feedback and Suggestion
Limitations of decision model
1 difficult to quantify safe and efficient operations
32
![Page 79: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/79.jpg)
Experts’ Feedback and Suggestion
Limitations of decision model
1 difficult to quantify safe and efficient operations
2 does not model individual human controllers
32
![Page 80: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/80.jpg)
Experts’ Feedback and Suggestion
Limitations of decision model
1 difficult to quantify safe and efficient operations
2 does not model individual human controllers
Suggestions for solution algorithm
32
![Page 81: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/81.jpg)
Experts’ Feedback and Suggestion
Limitations of decision model
1 difficult to quantify safe and efficient operations
2 does not model individual human controllers
Suggestions for solution algorithm
• return ∼ 3 good advisories
32
![Page 82: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/82.jpg)
Experts’ Feedback and Suggestion
Limitations of decision model
1 difficult to quantify safe and efficient operations
2 does not model individual human controllers
Suggestions for solution algorithm
• return ∼ 3 good advisories
• one advisory: optimal for decision model
32
![Page 83: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/83.jpg)
Experts’ Feedback and Suggestion
Limitations of decision model
1 difficult to quantify safe and efficient operations
2 does not model individual human controllers
Suggestions for solution algorithm
• return ∼ 3 good advisories
• one advisory: optimal for decision model
• other advisories: not too sub-optimal
32
![Page 84: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/84.jpg)
Experts’ Feedback and Suggestion
Limitations of decision model
1 difficult to quantify safe and efficient operations
2 does not model individual human controllers
Suggestions for solution algorithm
• return ∼ 3 good advisories
• one advisory: optimal for decision model
• other advisories: not too sub-optimal
• distinct advisories
32
![Page 85: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/85.jpg)
Advisory Difference Metric
Φ(C,C′)
Φ(C,C′) =
K∑
k=1
ϕ(Ck, C′k)
33
![Page 86: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/86.jpg)
Advisory Difference Metric
Φ(C,C′)
Φ(C,C′) =
K∑
k=1
ϕ(Ck, C′k)
Configuration difference metric
ϕ(Ck , C′k) =
¨
1 if Ck and C′kcombine sectors differently
0 else
33
![Page 87: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/87.jpg)
Multiple-Advisories Problem Statement
M ϵ-Optimal d-Distinct Configuration Schedule
Advisories Problem (M-ϵ-d-CSAs)
minimize
M∑
m=1
g(Cm, T)
subject to |CM| =M
Cmk∈ Ck k = 0,1,2, . . . , K, m = 1,2, . . . ,M
C1 ∈ C⋆
CSA( = optimal advisories for CSA)
g(Cm, T)− g(C1, T)
g(C1, T)≤ ϵ m = 2,3, . . . ,M
Φ(Cm, Cm′
) ≥ d ∀m,m′ where m 6=m′
34
![Page 88: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/88.jpg)
Multiple-Advisories Problem Statement
M ϵ-Optimal d-Distinct Configuration Schedule
Advisories Problem (M-ϵ-d-CSAs)
minimize
M∑
m=1
g(Cm, T)
subject to |CM| =M
Cmk∈ Ck k = 0,1,2, . . . , K, m = 1,2, . . . ,M
C1 ∈ C⋆
CSA( = optimal advisories for CSA)
g(Cm, T)− g(C1, T)
g(C1, T)≤ ϵ m = 2,3, . . . ,M
Φ(Cm, Cm′
) ≥ d ∀m,m′ where m 6=m′
M-ϵ-d-CSAs is NP-complete
34
![Page 89: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/89.jpg)
Algorithms for M-ϵ-d-CSAs
35
![Page 90: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/90.jpg)
Algorithms for M-ϵ-d-CSAs
1 Sequential Distinct A∗ with Shortcuts (SDA∗-SC)
• novel algorithm based on A∗
35
![Page 91: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/91.jpg)
Sequential Distinct A∗ (SDA∗)
Finding advisory C1:
C, T C1shortest path
algorithm
36
![Page 92: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/92.jpg)
Sequential Distinct A∗ (SDA∗)
Finding advisory Cm:
C, T Cm
{C1, C2, . . . , Cm−1}
constrained
shortest path
algorithm
36
![Page 93: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/93.jpg)
Sequential Distinct A∗ (SDA∗)
Finding advisory Cm:
C, T Cm
{C1, C2, . . . , Cm−1}
constrained
shortest path
algorithm
constrained shortest path problem is NP-complete
36
![Page 94: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/94.jpg)
Constrained Shortest Path Algorithm:
Forward Distinct A∗ (FDA∗)
For some λ ≥ 0, find C̃2 that minimizes Lagrangian
L(C2, λ) = g(C2, T)︸ ︷︷ ︸
advisorycost
+λ�
d− Φ(C1, C2)�
︸ ︷︷ ︸
similaritycost
37
![Page 95: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/95.jpg)
Constrained Shortest Path Algorithm:
Forward Distinct A∗ (FDA∗)
For some λ ≥ 0, find C̃2 that minimizes Lagrangian
L(C2, λ) = g(C2, T)︸ ︷︷ ︸
advisorycost
+λ�
d− Φ(C1, C2)�
︸ ︷︷ ︸
similaritycost
Another shortest path problem → use A∗
37
![Page 96: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/96.jpg)
FDA∗ and Duality
Proposition
If M = 2, ϵ =∞, and |C⋆CSA| = 1,
then FDA∗ implements the dual objective:
h(λ) =minimizeC2∈C
L(C2, λ)
38
![Page 97: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/97.jpg)
FDA∗ and Duality
Proposition
If M = 2, ϵ =∞, and |C⋆CSA| = 1,
then FDA∗ implements the dual objective:
h(λ) =minimizeC2∈C
L(C2, λ)
Corollary: If M = 2, ϵ =∞, |C⋆CSA| = 1,
λ = λ⋆, and strong duality holds,
then C̃2 satisfies a necessary condition for C2⋆
(C̃2 = second advisory returned by FDA∗)
38
![Page 98: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/98.jpg)
Sequential Distinct A∗ (SDA∗)
Finding advisory Cm:
C,T Cm
{C1,C2, . . . ,Cm−1}
constrained
shortest path
algorithm
constrained shortest path problem is NP-complete
39
![Page 99: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/99.jpg)
Sequential Distinct A∗ (SDA∗)
Finding advisory Cm:
C,T Cm
{C1,C2, . . . ,Cm−1}
FDA∗
λ
39
![Page 100: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/100.jpg)
Algorithms for M-ϵ-d-CSAs
1 Sequential Distinct A∗ with Shortcuts (SDA∗-SC)
• novel algorithm based on A∗
40
![Page 101: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/101.jpg)
Algorithms for M-ϵ-d-CSAs
1 Sequential Distinct A∗ with Shortcuts (SDA∗-SC)
• novel algorithm based on A∗
2 Forward Backward Value Iterationwith Sequential Advisory Search (FBVISAS)
• novel algorithm based on value iteration
40
![Page 102: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/102.jpg)
Forward Backward Value Iteration
with Sequential Advisory Search (FBVISAS)
0timestep1 2 3 4 5
41
![Page 103: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/103.jpg)
Forward Backward Value Iteration
with Sequential Advisory Search (FBVISAS)
0timestep1 2 3 4 5
41
![Page 104: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/104.jpg)
Forward Backward Value Iteration
with Sequential Advisory Search (FBVISAS)
0timestep1 2 3 4 5
41
![Page 105: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/105.jpg)
Forward Backward Value Iteration
with Sequential Advisory Search (FBVISAS)
0timestep1 2 3 4 5
41
![Page 106: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/106.jpg)
Theoretical Guarantees for FBVISAS:
the Good News
42
![Page 107: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/107.jpg)
Theoretical Guarantees for FBVISAS:
the Good News
Simple M-ϵ-d-CSAs instance: M = 2, ϵ =∞, |C⋆CSA| = 1,
C1 = C2 = · · · = CK, and C0 ∈ C1
42
![Page 108: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/108.jpg)
Theoretical Guarantees for FBVISAS:
the Good News
Simple M-ϵ-d-CSAs instance: M = 2, ϵ =∞, |C⋆CSA| = 1,
C1 = C2 = · · · = CK, and C0 ∈ C1
Reconfiguration cost-dominated M-ϵ-d-CSAs instance:
reconfiguring never decreases the advisory cost
42
![Page 109: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/109.jpg)
Theoretical Guarantees for FBVISAS:
the Good News
Simple M-ϵ-d-CSAs instance: M = 2, ϵ =∞, |C⋆CSA| = 1,
C1 = C2 = · · · = CK, and C0 ∈ C1
Reconfiguration cost-dominated M-ϵ-d-CSAs instance:
reconfiguring never decreases the advisory cost
Proposition
If one exists, FBVISAS finds an optimal C2⋆ for
simple reconfiguration cost-dominated instances
42
![Page 110: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/110.jpg)
Theoretical Guarantees for FBVISAS:
the Bad News
Static cost-dominated M-ϵ-d-CSAs instance: βR = 0
43
![Page 111: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/111.jpg)
Theoretical Guarantees for FBVISAS:
the Bad News
Static cost-dominated M-ϵ-d-CSAs instance: βR = 0
Proposition
If d > 1, FBVISAS does not return any C2 for
simple static cost-dominated instances
43
![Page 112: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/112.jpg)
Algorithms for M-ϵ-d-CSAs
1 Sequential Distinct A∗ with Shortcuts (SDA∗-SC)
• novel algorithm based on A∗
2 Forward Backward Value Iterationwith Sequential Advisory Search (FBVISAS)
• novel algorithm based on value iteration
44
![Page 113: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/113.jpg)
Algorithms for M-ϵ-d-CSAs
1 Sequential Distinct A∗ with Shortcuts (SDA∗-SC)
• novel algorithm based on A∗
2 Forward Backward Value Iterationwith Sequential Advisory Search (FBVISAS)
• novel algorithm based on value iteration
3 Lowest-Cost Paths (LCP)
• finds optimal solution to relaxation• many efficient algorithms
44
![Page 114: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/114.jpg)
Algorithms for M-ϵ-d-CSAs
1 Sequential Distinct A∗ with Shortcuts (SDA∗-SC)
• novel algorithm based on A∗
2 Forward Backward Value Iterationwith Sequential Advisory Search (FBVISAS)
• novel algorithm based on value iteration
3 Lowest-Cost Paths (LCP)
• finds optimal solution to relaxation• many efficient algorithms
4 Value Iteration Fraction Optimalwith Exhaustive Advisory Search
• finds optimal solution• not computationally efficient
44
![Page 115: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/115.jpg)
Modeling Decisions and Developing Algorithms
fast-time
simulations
45
![Page 116: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/116.jpg)
Modeling Decisions and Developing Algorithms
fast-time
simulations
45
![Page 117: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/117.jpg)
Evaluation of Algorithms on Small Instances
47
48
49
4546 4546
4749
48
• traffic and configurations from
two days in December 2011
• nine two-hour instances per day
• 18 total instances
• five-minute time steps
• C requires same number of open
sectors as were used operationally
• request two advisories (M = 2)
• ϵ = 0.2 constraint on cost of C2
• d requires 30 minutes of different
airspace configurations
46
![Page 118: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/118.jpg)
Small Instances: Properties of C2
SDA∗-SC FBVISAS LCP0
20
40
60
80
100
Percent ofInstances
optimal C2⋆
47
![Page 119: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/119.jpg)
Small Instances: Properties of C2
SDA∗-SC FBVISAS LCP0
20
40
60
80
100
Percent ofInstances
optimal C2⋆
feasible C2
47
![Page 120: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/120.jpg)
Evaluation of Algorithms on Realistic Instances
47
48
49
4546 4546
4749
48
• traffic and configurations from
231 days in 2011 and 2012
• 6 am to midnight local time
• rolling horizon: implement first
hour of two-hour advisories
• 4158 total instances
• five-minute time steps
• C only requires same initial
configuration as was used
operationally
• request three advisories (M = 3)
• ϵ = 0.5 constraint on cost
• d requires 30 minutes of different
airspace configurations
48
![Page 121: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/121.jpg)
Realistic Instances: Advisory Costs
0.7 0.8 0.9 1.1 1.2 1.30
20
40
60
80
100
0.99 1.01
Percent ofInstances
Cost Ratio�g(C2 ,T)+g(C3 ,T) for SDA∗-SC
g(C2 ,T)+g(C3 ,T) for FBVISAS
�
SDA∗-SCworse
SDA∗-SCbetter
mean = 1.02
49
![Page 122: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/122.jpg)
Realistic Instances: Computation Times
(on a Workstation)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Computation
Times
[seconds]
SDA∗-SC FBVISAS
50
![Page 123: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/123.jpg)
Modeling Decisions and Developing Algorithms
fast-time
simulations
51
![Page 124: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/124.jpg)
Modeling Decisions and Developing Algorithms
51
![Page 125: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/125.jpg)
Operational Airspace Sectorization
Integrated System (OASIS)
52
![Page 126: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/126.jpg)
OASIS Screenshot: Multiple Advisories
53
![Page 127: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/127.jpg)
Human-in-the-Loop Simulations Set Up
• eight retired FAA
personnel
• four simulated
scenarios
54
![Page 128: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/128.jpg)
Human-in-the-Loop Simulations Set Up
• eight retired FAA
personnel
• four simulated
scenarios
Three experimental conditions in which users
1 generate configuration schedule
2 select from among algorithm-generated advisories
3 select from among algorithm-generated advisories
and then modify
54
![Page 129: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/129.jpg)
Human-in-the-Loop Simulations Results
[Lee et al., 2013]
55
![Page 130: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/130.jpg)
Human-in-the-Loop Simulations Results
[Lee et al., 2013]
Advisories enable safe and efficient operations
• average acceptability of selected
algorithm-generated advisories: > 4 out of 5
• user modifications were minor and led to
no significant improvement in acceptability
55
![Page 131: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/131.jpg)
Human-in-the-Loop Simulations Results
[Lee et al., 2013]
Advisories enable safe and efficient operations
• average acceptability of selected
algorithm-generated advisories: > 4 out of 5
• user modifications were minor and led to
no significant improvement in acceptability
Providing multiple advisories added value
• in more than 60% of instances, users selected
second or third advisory
• when asked how many advisories they wanted the
tool to provide, users requested an average of 2.8
55
![Page 132: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/132.jpg)
Human-in-the-Loop Simulations Results
[Lee et al., 2013]
Advisories enable safe and efficient operations
• average acceptability of selected
algorithm-generated advisories: > 4 out of 5
• user modifications were minor and led to
no significant improvement in acceptability
Providing multiple advisories added value
• in more than 60% of instances, users selected
second or third advisory
• when asked how many advisories they wanted the
tool to provide, users requested an average of 2.8
Computation times were rated as highly acceptable
55
![Page 133: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/133.jpg)
Modeling Decisions and Developing Algorithms
56
![Page 134: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/134.jpg)
Outline
• Air traffic management background
• Research topics
• Research objective and approach
• Case study:
decision-support tool for area supervisors
• Summary of contributions
57
![Page 135: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/135.jpg)
Summary of Contributions
58
![Page 136: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/136.jpg)
Summary of Contributions
1 Area supervisors:area configuration selection
– developed prescriptive decision model– designed efficient algorithms to findlow-cost and distinct paths
– implemented algorithm in decision-support tool
58
![Page 137: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/137.jpg)
Summary of Contributions
1 Area supervisors:area configuration selection
– developed prescriptive decision model– designed efficient algorithms to findlow-cost and distinct paths
– implemented algorithm in decision-support tool
2 Operations managers:assignment of flights to slots
– designed maximum-likelihood-based algorithm that– generated insights from operational decision data
58
![Page 138: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/138.jpg)
Summary of Contributions
1 Area supervisors:area configuration selection
– developed prescriptive decision model– designed efficient algorithms to findlow-cost and distinct paths
– implemented algorithm in decision-support tool
2 Operations managers:assignment of flights to slots
– designed maximum-likelihood-based algorithm that– generated insights from operational decision data
3 Flow managers:Ground Delay Program implementation
– deployed supervised learning and inversereinforcement learning models that
– produced predictive capability and insights
58
![Page 139: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/139.jpg)
Relevant peer-reviewed conference papers1 M. Bloem and P. Gupta, "Configuring Airspace Sectors with Approximate Dynamic
Programming," ICAS, September 2010.
2 M. Bloem and H. Huang, "Evaluating Delay Cost Functions with Airline Actions inAirspace Flow Programs," USA/Europe ATM R&D Seminar, June 2011.
3 M. Bloem and N. Bambos, "Coordinated Tactical Air Traffic and Airspace Management,"IEEE CDC, December 2011.
4 M. Bloem, M. Drew, C. F. Lai, and K. Bilimoria, "Advisory Algorithm for Scheduling OpenSectors, Operating Positions, and Workstations," AIAA ATIO, September 2012.
5 M. Bloem, H. Huang, and N. Bambos, "Approximating the Likelihood of Historical AirlineActions to Evaluate Airline Delay Cost Functions," IEEE CDC, December 2012.
6 M. Bloem and N. Bambos, "An Approach for Finding Multiple Area of SpecializationConfiguration Advisories," AIAA ATIO, August 2013.
7 M. Bloem and N. Bambos, "Ground Delay Program Analytics with Behavioral Cloningand Inverse Reinforcement Learning," AIAA ATIO, June 2014.
8 M. Bloem and N. Bambos, "Infinite Time Horizon Maximum Causal Entropy InverseReinforcement Learning," IEEE CDC, December 2014.
Relevant journal articles1 M. Bloem, P. Gupta, and P. Kopardekar, "Algorithms for Combining Airspace Sectors," Air
Traffic Control Quarterly, Vol. 17, No. 4, 2009.
2 M. Bloem, M. Drew, C. F. Lai, and K. D. Bilimoria, "Advisory Algorithm for Scheduling OpenSectors, Operating Positions, and Workstations," AIAA Journal of Guidance, Control, andDynamics, Vol. 37, No. 4, July–August 2014.
3 M. Bloem and N. Bambos, "Air Traffic Control Configuration Advisories fromNear-Optimal Distinct Paths," AIAA Journal of Aerospace Information Systems, Vol. 11,No. 11, November 2014.
4 M. Bloem and N. Bambos, "Ground Delay Program Analytics with Behavioral Cloningand Inverse Reinforcement Learning," AIAA Journal of Aerospace Information Systems,accepted on 21 December 2014, available online on 03 March 2015.
5 M. Bloem and N. Bambos, "Stochastic Models of Ground Delay ProgramImplementation for Prediction, Simulation, and Insight," Journal of AerospaceOperations, invited submission to special issue, expected publication in late 2015.
59
![Page 140: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/140.jpg)
Acknowledgments
• Stanford
• NASA
• Family & friends
60
![Page 141: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/141.jpg)
Questions?
![Page 142: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/142.jpg)
Backup Slides
62
![Page 143: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/143.jpg)
Why Improve Area Configuration Planning?
1 more efficient flights
2 fewer disruptions to area supervisors and
traffic managers
3 better controller staff management
4 prepare for increased flexibility in future operations
→ more complex planning
63
![Page 144: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/144.jpg)
Modeling Decisions and Developing Algorithms
fast-time
simulations
64
![Page 145: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/145.jpg)
Operational Decision Data Set and
Fast-Time Simulation Setup
47
48
49
4546 4546
4749
48
• traffic and configurations from
230 days in 2011 and 2012
• 6 am to midnight local time
• rolling horizon: implement first
hour of two-hour advisories
• five-minute time steps
• only airspace configurations
• constraint: only use typical
configurations
65
![Page 146: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/146.jpg)
Static Cost versus Reconfiguration Cost
700 800 900 1000 11000
50
100
150
200
Total
Reconfiguration
Cost
Total Static Cost
operational
βR = 0.5
βR = 15
ratio(βR) =total reconfiguration cost(βR)
total static cost(βR)
error(βR) =��ratio(operational)− ratio(βR)
��
66
![Page 147: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/147.jpg)
Selecting βR
0 1.75 5.0 10 155
10
15
20
25
Total Error
βR
set βR = 1.75
67
![Page 148: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/148.jpg)
Uncertainty in Traffic Predictions:
Why Not Incorporated?
• Experts understand prediction errors
• Avoid inconsistency with other deterministic tools
68
![Page 149: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/149.jpg)
Uncertainty in Traffic Predictions:
Uncertain Future Aircraft Counts
0 5 10 15 200
0.05
0.1
0.15
0.2
0.25
Aircraft Count
Probability
prediction
distribution
69
![Page 150: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/150.jpg)
Uncertainty in Traffic Predictions:
Impact on Advisory Costs
Advisories generated using predicted traffic with realistic errors
are typically 2.5%–12.5% costlier
than advisories generated with perfectly-predicted traffic
70
![Page 151: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/151.jpg)
Uncertainty in Traffic Predictions:
Approximate Dynamic Programming Approach
• Finite time horizon Markov decision process (MDP)
– state: sector aircraft counts and predictions thereof– action: current configuration (not a schedule)– state transitions: based on prediction errors inoperational tool
– objective: similar to decision model (CSA)
• Rollouts algorithm performs 15% better than a
heuristic and within 2% of optimal
• Rollouts algorithm computes solutions fast enough
for real-time implementation
M. Bloem and P. Gupta, "Configuring Airspace Sectors withApproximate Dynamic Programming," ICAS, September 2010.
71
![Page 152: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/152.jpg)
Uncertainty in Traffic Predictions:
Other Approaches
• improve predictions with a statistical model
• discount contribution to cost by predicted aircraft
• stochastic shortest path problem
• risk-averse advisory
• larger MDP formulation
• robust set of advisories
72
![Page 153: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/153.jpg)
SDA∗ with Shortcuts (SDA∗-SC)
Objectives:
1 computation time ≤ that of A∗ for each advisory
→ use data from reverse A∗ for C1
to find "shortcuts"
2 no re-tuning of λ per instance or advisory
→ normalize advisory cost and similarity cost terms
73
![Page 154: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/154.jpg)
Forward Backward Value Iteration
with Sequential Advisory Search (FBVISAS)
• Use value iteration to find minimum-cost advisory
through each Ck ∈ Ck, k = 1, . . . ,K
• Sort these advisories from low to high cost
• Initialize set of advisories to return: CM ← ∅
• Loop:
– select next advisory– if advisory is sufficiently different from otheradvisories in C
M, add it to CM
– if |CM| =M, stop
74
![Page 155: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/155.jpg)
Multiple Advisories Problem is NP-Complete
For an arbitrary instance of the independent set
problem on G(V,E), construct M-ϵ-d-CSAs instance:
• M = required number of independent nodes
• gk() = 0
• K = 1 and C1 = V
• d = 1
• configuration difference metric:
ϕ(Ck,C′k) =
¨
1 if Ck and C′kindependent in G(V,E)
0 else
∃ solution to the independent set instance
⇔ ∃ solution to the M-ϵ-d-CSAs instance
75
![Page 156: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/156.jpg)
Computational Complexity
of Multiple Solutions Problem Algorithms
Algorithm Complexity
VIFOEAS O(n3K)a
FBVISAS O(n2K + nK log(nK + 1))
SDA∗-SC O(Mn2K log(nK +1))
Eppstein lowest-cost paths O(n2K + nK log(nK + 1) +M)
Suurballe lowest-cost node-disjoint paths O(Mn2K log(nK +1))b
a This does not include the complexity of searching through�|Cϵ |
M
�advisory
subsets.b This assumes that Dijkstra’s algorithm is used as a subroutine for findingshortest paths.
76
![Page 157: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/157.jpg)
Airline Delay Cost Model Evaluation
decision-
support
tool
77
![Page 158: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/158.jpg)
Airline Delay Cost Model Evaluation
Motivation: how costly is a flight delay?
78
![Page 159: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/159.jpg)
Airline Delay Cost Model Evaluation
Motivation: how costly is a flight delay?
Operational decision data:
default assignment
ABC1
10:00
ABC2
10:30
Slot #60
11:00
Slot #90
12:00
78
![Page 160: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/160.jpg)
Airline Delay Cost Model Evaluation
Motivation: how costly is a flight delay?
Operational decision data:
airline-selected assignment
ABC1
10:00
ABC2
10:30
Slot #60
11:00
Slot #90
12:00
78
![Page 161: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/161.jpg)
Airline Delay Cost Model Evaluation
Motivation: how costly is a flight delay?
Operational decision data:
airline-selected assignment
ABC1
10:00
ABC2
10:30
Slot #60
11:00
Slot #90
12:00
Contribution: novel algorithm enables determination of
delay cost model and cost noise parameters that
maximize an approximation of likelihood of data
78
![Page 162: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/162.jpg)
Airline Delay Cost Model Evaluation
• developed novel algorithm for finding cost noise
parameters that maximize an approximation of the
likelihood of minimum-cost matching problem
instance solutions, given a candidate cost model
79
![Page 163: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/163.jpg)
Airline Delay Cost Model Evaluation
• developed novel algorithm for finding cost noise
parameters that maximize an approximation of the
likelihood of minimum-cost matching problem
instance solutions, given a candidate cost model
• evaluated more than ten proposed airline delay
cost models using hundreds of slot swap decisions
79
![Page 164: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/164.jpg)
Airline Delay Cost Model Evaluation
• developed novel algorithm for finding cost noise
parameters that maximize an approximation of the
likelihood of minimum-cost matching problem
instance solutions, given a candidate cost model
• evaluated more than ten proposed airline delay
cost models using hundreds of slot swap decisions
• determined delay cost model that maximizes
approximate likelihood of data for each airline
79
![Page 165: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/165.jpg)
Airline Delay Cost Model Evaluation
• developed novel algorithm for finding cost noise
parameters that maximize an approximation of the
likelihood of minimum-cost matching problem
instance solutions, given a candidate cost model
• evaluated more than ten proposed airline delay
cost models using hundreds of slot swap decisions
• determined delay cost model that maximizes
approximate likelihood of data for each airline
• estimated cost noise parameters for each delay
cost model for each airline
79
![Page 166: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/166.jpg)
Ground Delay Program (GDP)
Implementation Modeling
decision-
support
tool
80
![Page 167: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/167.jpg)
Ground Delay Program (GDP)
Implementation Modeling
Operational decision data:
arrival rate control arrivals
start end per
airport time time hour
SFO 9:00 am noon 30
81
![Page 168: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/168.jpg)
Ground Delay Program (GDP)
Implementation Modeling
Operational decision data:
arrival rate control arrivals
start end per
airport time time hour
SFO 9:00 am noon 30
Motivation: predict & understand GDP implementation
81
![Page 169: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/169.jpg)
Ground Delay Program (GDP)
Implementation Modeling
Operational decision data:
arrival rate control arrivals
start end per
airport time time hour
SFO 9:00 am noon 30
Motivation: predict & understand GDP implementation
Contribution: produced predictive models and insights
by developing behavioral cloning and inverse
reinforcement learning models of GDP implementation
81
![Page 170: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/170.jpg)
Ground Delay Program (GDP)
Implementation Modeling
• evaluated behavioral cloning and inverse
reinforcement learning models of GDP
implementation at EWR and SFO using hundreds of
days of operational decision data
82
![Page 171: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/171.jpg)
Ground Delay Program (GDP)
Implementation Modeling
• evaluated behavioral cloning and inverse
reinforcement learning models of GDP
implementation at EWR and SFO using hundreds of
days of operational decision data
• demonstrated that a behavioral cloning model can
produce superior predictions of GDP
implementation
82
![Page 172: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/172.jpg)
Ground Delay Program (GDP)
Implementation Modeling
• evaluated behavioral cloning and inverse
reinforcement learning models of GDP
implementation at EWR and SFO using hundreds of
days of operational decision data
• demonstrated that a behavioral cloning model can
produce superior predictions of GDP
implementation
• determined that neither class of model provides
much evidence that conditions beyond those in the
next two hours impact GDP implementation
82
![Page 173: Michael Bloem PhD Defense - Optimization and Analytics for Air Traffic Management](https://reader033.fdocuments.us/reader033/viewer/2022042717/55d6f3e1bb61eb075e8b4603/html5/thumbnails/173.jpg)
Ground Delay Program (GDP)
Implementation Modeling
• evaluated behavioral cloning and inverse
reinforcement learning models of GDP
implementation at EWR and SFO using hundreds of
days of operational decision data
• demonstrated that a behavioral cloning model can
produce superior predictions of GDP
implementation
• determined that neither class of model provides
much evidence that conditions beyond those in the
next two hours impact GDP implementation
• estimated a reward function that provides insight
into metrics guiding GDP implementation
82