A new method for Conflict Detection and Resolution in

cathamΤεχνίτη Νοημοσύνη και Ρομποτική

23 Οκτ 2013 (πριν από 3 χρόνια και 10 μήνες)

224 εμφανίσεις

Semantic Cities @ AAAI 2012

July 23rd, 2012

University of Tabriz

Faculty of Electrical and Computer Engineering

In The Name of God

A new method for Conflict Detection and Resolution in
Air Traffic Management

By:

Farnaz Derakhshan

Hojjat Emami

Contents

2
-
28

References

7

Conflict Detection and Resolution Process



2

Imperialist Competitive Algorithm (ICA)

4

Content

Introduction

1

Semantic Cities @ AAAI 2012

Our Proposed Model

5

Test Results & Conclusion

6

Graph Coloring Problem (GCP)

3

Introduction (Air Traffic Control)


Air traffic:


“Aircraft operating in the air or on an airport surface, exclusive of loading ramps
and parking areas”

(Federal Aviation Regulations and Aeronautical Information Manual, 2010 edition)



Air Traffic Control:


A service operated by appropriate authority to promote the safe, orderly, and
expeditious flow of air traffic

(Federal Aviation Regulations and Aeronautical Information Manual, 2010 edition)

Introduction

Semantic Cities @ AAAI 2012

3
-
28

Introduction (Air Traffic Control)


Air traffic control is a complicated task, involving multiple and dynamic controls and
have high level of granularity.



having a reliable, safe and efficient air traffic management is a fundamental and
critical need in aviation industry.


Introduction

Semantic Cities @ AAAI 2012

4
-
28

Introduction (Free Flight)


Free flight is a new concept presented potentially to solve problems in the current air
traffic management system.


Free flight method has many advantages (such as less fuel consumption, minimum
delays, reduction of the workload of the air traffic control centers, high efficiency,
and has distributed nature)



The most notable problem in free flight method is the
occurrence of conflicts

between different aircrafts’ flights.



Introduction

5
-
28

Semantic Cities @ AAAI 2012

Introduction


One of the fundamental challenges in the current air traffic management and
especially in the free flight is detection and resolution of conflicts.


We presented a new model for conflict detection and resolution between aircrafts in
airspace using graph coloring problem.


We mapped the conflict detection and resolution problem to graph coloring problem.


To solve graph coloring problem we used imperialist competitive algorithm.



Introduction

6
-
28

Semantic Cities @ AAAI 2012

Conflict Detection and Resolution Process


Conflict
:


Conflict

is

the

event

in

which

two

or

more

than

two

aircrafts

experience

a

loss

of

minimum

separation

from

each

other


Conflict

Detection
:



The

process

of

deciding

when

conflict

(conflict

between

aircrafts)

will

occur


Conflict

Resolution
:


specifying

what

action

and

how

should

be

to

resolve

conflicts

Conflict Detection and Resolution Process

7
-
28

Semantic Cities @ AAAI 2012

Conflict Detection and Resolution (General view)

Conflict Detection and Resolution (General view)

8
-
28

No

Yes

Conflict Resolution

Conflict
Detected?

Environment (airspace)

Conflict Detection

Operating area
-

Traffic
Information


weather

conditions and …




Operator
or

Agent






Semantic Cities @ AAAI
2012

Figure.
1
: Conflict Detection and
Resolution (General view)

Graph Coloring Problem (GCP)


GCP is an optimization problem which finds an optimal coloring for a given graph G.








(Jensen and
Toft

1995
)



GCP is one of the NP
-
hard problems


GCP is a practical method of representing many real world problems including:


Time scheduling


Frequency assignment


Register allocation


Circuit board testing




Graph Coloring Problem (GCP)

9
-
28

Semantic Cities @ AAAI
2012

Graph Coloring Problem (Cont)


Given

an

undirected

graph

G=(V,

E)

with

a

set

of

vertices

V

and

a

set

of

edges

E
;


A

K
-
coloring

of

G

includes

assigning

a

color

to

each

vertex

of

V,

such

that

neighboring

vertices

have

different

colors

(labels)
.


GCP

implemented

by

using

a

conflict

minimization

algorithm

Graph Coloring Problem (Cont)

10
-
28

Semantic Cities @ AAAI
2012

1

2

3

4

1

2

3

4

An Undirected Graph before Coloring

|V| =
4

|E| =
4

Chromatic number =
2



|K| =
2
(Colors : Red, Green)

Colored Graph

Figure.
2
: an example of coloring a graph

Search

Strategies

Search Strategies



Complete

Deterministic

Evolutionary Programming

Evolutionary Strategy

Genetic Algorithms

Genetic Programming

Heuristic

(
Non
-
Deterministic
)

Estimation
of Distribution
Algorithm

Evolutionary
Algorithms

Population
-
based

Search Strategies

Imperialist Competitive Algorithm

Point
-
based

11
-
28

Semantic Cities @ AAAI
2012

Imperialist Competitive Algorithm (ICA)


Imperialist

Competitive

Algorithm

(ICA)

is

Novel

Socio
-
politically

Motivated

Optimization

Strategy
.


Proposed

by

Atashpaz
-
Gargari

and

Lucas,

2007
.



is

inspired

by

sociopolitical

process

of

imperialism

!!


since

in

late

inception,

it

has

been

used

in

many

applications
.


has

shown

good

convergence

and

global

minimum

achievement
.


has

a

lot

to

do

with
.



Introduction(ICA)

12
-
28

Semantic Cities @ AAAI
2012

A Big Picture of ICA

A Big Picture

Figure.
3
: A big picture of ICA (
Gargari

and Lucas,
2007
)

13
-
28

Semantic Cities @ AAAI
2012

Flowchart of the ICA

Flowchart of the ICA




start

Is there an empire


With no colonies

Yes

Done

Compute the total cost of all empires

No

No

is there

a

colony in an empire

which

has lower cost

than that of

imperialist


Exchange the positions of that


Imperialist & colony

Initialize the
Emp
ire
s

pick the weakest colony from the


weakest empire
and

give
it to

the empire that has the most


likelihood to posses it

Eliminate this

empire

Yes

move the colonies


toward the imperialist

Yes

stop condition


satisfied

*

*

No

End

14
-
28

Semantic Cities @ AAAI
2012

Figure.
4
: Flowchart of the ICA

Our Proposed Model


In our model, the main strategy is based on: "Prevention is better than cure“


we mapped the conflict detection and resolution problem to graph coloring problem.


we map the aircrafts congestion area to a corresponding graph based on aircrafts
condition in airspace.


imperialist competitive algorithm has shown great performance in both convergence
rate and better global optimal achievement, therefore we used this algorithm to solve
graph coloring problem rather than other evolutionary algorithms.



In our model, a Global approach is used for resolving the multiple conflicts.

(the entire traffic situation is examined simultaneously, and it has the high capabilities)

Our Proposed Model

15
-
28

Semantic Cities @ AAAI
2012

Our Proposed Model



start

Solving the GCP using ICA

(Conflict Resolution process)

Map the Congestion Area to a Graph

(Making Adjacency Matrix)

Monitor the Environment

Project current

states to future states

by using nominal propagation method

Detect the Congestion Area

Compute Distance between all Aircrafts

in Congestion Area

Define Problem Parameters

And other Traffic Parameters

16
-
28

Send new Flight Paths Plan to Existing

Aircrafts in Congestion Area

End

Figure.
5
: Block diagram of our proposed model

Semantic Cities @ AAAI
2012

Our Proposed Model

Creating the Graph


here,

the

nodes

of

graph

indicate

aircrafts

in

the

congestion

area

and

every

edge

between

two

nodes

indicates

a

probable

conflicts

in

the

future

time

step,

if

the

aircrafts

continue

their

current

flight

plan
.

Creating the Graph

17
-
28

Semantic Cities @ AAAI
2012

Figure.
6
: Graphical display of an example

conflict detection & resolution scenario

(creating the graph)

Conflict Detection


we

use

a

simple

conflict

prediction

method
.


we

used

nominal

state

propagation

method

for

prediction

conflicts

between

aircrafts
.


the

current

position,

heading

and

speed

of

existing

aircrafts

in

congestion

area

used

for

mapping

the

current

location

to

the

future

states
.



in

the

future

states

if

distance

between

two

aircraft

is

less

than

a

predefined

reliable

distance

threshold

we

say

a

conflict

is

going

to

occur
.


(i.e. when in future states the protected zones of two aircraft is overlapped then
system report a conflict)



here,

we

focused

only

horizontal

plan

dimension


Conflict Detection

18
-
28

Semantic Cities @ AAAI
2012

Conflict Resolution


we

map

the

congestion

area

to

a

state

space

graph
.


we

replace

solving

the

conflicts

between

different

aircrafts

with

coloring

the

corresponding

graph
.


we

used

imperialist

competitive

algorithm

to

solve

graph

coloring

problem


a

colored

plan

for

graph

is

a

reliable

solution

for

conflicts

problem


in

the

conflict

resolution

process,

for

each

aircraft,

we

allocate

a

flight

path

in

which

this

aircraft

will

has

a

reliable

distance

with

other

aircrafts

and

there

is

no

risk

of

conflict
.


In

our

model,

we

can

assume

each

flight

path

as

five

directional

options

namely
:

main

line,

deviation

to

right

of

the

main

line,

deviation

to

left

of

the

main

line,

top

of

the

main

line

and

bottom

of

the

main

line
.




Conflict Resolution

19
-
28

Semantic Cities @ AAAI
2012

Test Results


To

evaluate

our

proposed

model,

we

use

random

flights

model

(Archibald

et

al
.

2008
)


here

all

aircrafts

are

constrained

to

fly

at

the

same

altitude

and

at

a

constant

speed
.



Small

and

instantaneous

heading

changes

for

each

aircraft

are

the

only

maneuvers

of

resolving

conflicts
.



We

used

supposed

scenarios

that

these

samples

contain

2
,

3
,

4
,

5
,

6
,

8
,

12
,

16
,

20

and

30

aircrafts

in

congestion

area
.


In

air

traffic

control

we

deal

with

a

multi

objective

problem
.


In

this

paper,

our

goal

is

providing

a

safe,

reliable

and

efficient

strategy

for

solving

conflicts

between

aircrafts
.


Test Results

20
-
28

Semantic Cities @ AAAI
2012

Performance Metrics


The

ideal

state

in

conflict

resolution

model

is

that

the

aircrafts

are

able

to

track

their

destination

without

deviation

or

with

minimal

deviation

from

their

original

path
.


Maneuvers

that

are

used

in

the

conflict

resolution

methods,

causes

the

aircraft

to

be

diverted

from

the

ideal

and

optimal

mainstream
.



System

efficiency

metric
:


measures

the

degree

to

which

the

aircrafts

in

the

system

are

able

to

follow

direct

and

linear

flight

paths

to

their

destinations
.



In

this

paper,

we

define

a

performance

criterion

for

each

aircraft

same

as

follows
:

Performance Metrics

21
-
28

Semantic Cities @ AAAI
2012

Performance Metrics (cont)


here,

we

define

a

performance

criterion

for

each

aircraft

same

as

follows
:










We try to select routes with lowest cost when we redirect the aircrafts’ main routes
(i.e. the lowest deviation from the main flight path for each aircraft)

Performance Metrics (cont)

22
-
28

ideal and optimal flight path for an aircraft

the amount of deviation from mainstream of aircrafts

Semantic Cities @ AAAI
2012

Performance Metrics (cont)


System

Efficiency
:








how

much

this

criterion

is

closer

to


1


indicates

good

performance

of

the

system

and

how

much

this

criterion

is

closer

to


0


indicates

poor

performance



our

proposed

model

for

solving

GCP

for

higher

dimensions

(for

a

great

number

of

aircrafts

that

are

in

the

congestion

area)

acts

as

a

good

way
.

Performance Metrics (cont)

23
-
28

Total number of aircrafts

Performance of each aircraft

Semantic Cities @ AAAI
2012

Test Results

Table
1
: Test results of applying the algorithm onto input graphs (congestion areas) with
specified parameters


Test Results

24
-
28

Semantic Cities @ AAAI
2012

Conclusion


one of the fundamental challenges in the current air traffic management and
especially in free flight method is detection and resolution of conflicts.


in our approach, we mapped conflict detection and resolution problem to graph
coloring problem, then use imperialist competitive algorithm to solve GCP.


Our proposed model provides an efficient and reliable solution for solving conflicts
in air traffic management


in conflict resolution process we tried to select routes with lowest cost when the
aircrafts’ redirected from main route, therefore, we have least delay in flights and
the minimum consumption of resources (e.g. in fuel consumption).


our proposed model use multiple strategies for the resolution of conflicts.

Conclusion

25
-
28

Semantic Cities @ AAAI
2012

References

1.
Federal

Aviation

Regulations

and

Aeronautical

Information

Manual,

2010

edition
.

2009

ASA,

Inc
.

Newcastle,

Washington
.

2.
Rong,

J
.
,

Geng
,

Sh
.
,

Valasek,

J
.
,

R
.

Ioerger,

T
.

2002
.

Air

traffic

conflict

negotiation

and

resolution

using

an

onboard

multi

agent

system
.

Digital

Avionics

systems

Conf
.

Irvine,

CA
.

3.
Kuchar,

J
.
,

and

Yang,

L
.

C
.

2000
.

A

Review

of

CD&R

Modeling

Methods
.

IEEE

Transactions

on

Intelligent

Transportation

Systems,

Vol
.

1
,

No
.

4
,

December
.

4.
Agogino,

A
.
,

and

Tumer
,

K
.

2008
.

Adaptive

Management

of

Air

Traffic

Flow
:

A

Multiagent

Coordination

Approach
.

Proceedings

of

the

Twenty
-
Third

AAAI

Conf
.

on

Artificial

Intelligence
.

5.
Agogino,

A
.
,

and

Tumer
,

K
.

2009
.

Improving

air

traffic

management

with

a

learning

multiagent

system
.

IEEE

Intell
.

Syst
.
,

vol
.

24
,

no
.

1
,

pp
.

18

21
,

Jan/Feb
.

6.
Atashpaz
-
Gargari
,

E
.
,

Lucas,

C
.

2007
.

Imperialist

Competitive

Algorithm
:

An

algorithm

for

optimization

inspired

by

imperialistic

competition
.

IEEE

Congress

on

Evolutionary

Computation,

4661

4667
.

7.
Eby
,

M
.

S
.

and

III,

W
.

E
.

K
.

1999
.

Free

flight

separation

assurance

using

distributed

algorithms
.

In

Proceedings

of

aerospace

conference
.

Aspen,

Co
.

8.
Archibald

k
.
,

Hill

c
.
,

Jepsen

a
.
,

Stirling

c
.
,

and

Frost

l
.

2008
.

A

satisfying

approach

to

aircraft

conflict

resolution
.

IEEE

transactions

on

systems,

man,

and

cybernetics

part

c
:

applications

and

reviews,

vol
.

38
,

no
.


References

26
-
28

Semantic Cities @ AAAI
2012

References

9.
Agogino,

A
.

K
.
,

and

Tumer
,

K
.

2009
.

Learning

indirect

actions

in

complex

domains
:

Action

suggestions

for

air

traffic

control
.

Advances

in

Complex

Systems,

12
,

493

512
.

10.
Whalen,

S
.

2002
.

Graph

Coloring

with

Genetic

Algorithms
.

http
:
//mat
.
gsia
.
cmu
.
edu/COLOR/color
.
html
.

11.
Garey
,

M
.
R
.

and

Johnson,

D
.
S
.

1999
.

Computers

and

intractability
:

a

guide

to

the

theory

of

NP
-
completeness,

W
.
H
.

Freeman

and

Company,

New

York
.

12.
De

Werra

D
.

1990
.

Heuristics

for

Graph

Coloring
.

Computing

Suppl
.

7
,

pp
.

191
-
208
.

13.
Goldberg,

D
.

E
.

1980
.

Genetic

Algorithms

in

Search,

Optimization

and

Machine

Learning
.

Addison
-
Wesley,

Reading,

MA
.

14.
Valkanas,

G
.
,

Natsiavas,P
.
,

Bassiliades,N
.

2009
.

A

collision

detection

and

resolution

multi

agent

approach

using

utility

functions,

Fourth

Balkan

Conf
.

in

Informatics
.

15.
Chao,

W
.

and

Jing,

G
.

2009
.

Analysis

of

Air

Traffic

Flow

Control

through

Agent
-
Based

Modelling

and

Simulation
.

International

Conf
.

on

Computer

Modelling

and

Simulation,

IEEE
.


16.
Jensen

T
.
R
.
,

Toft

B
.

1995
.

Graph

Coloring

Problems,

Wiley

interscience

Series

in

Discrete

Mathematics

and

Optimization
.

References

27
-
28

Semantic Cities @ AAAI
2012

The End & Thanks

The End & Thanks


T
hanks for your attention
!