semantic cities @ aaai 2012 july 23rd, 2012

Semantic Cities @ AAAI 2012

July 23rd, 2012

University of TabrizFaculty 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

References7

Conflict Detection and Resolution Process

2

Imperialist Competitive Algorithm (ICA)4

Content

Introduction 1


Our Proposed Model5

Test Results & Conclusion6

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 5-28Semantic 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 6-28Semantic Cities @ AAAI 2012


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


7-28Semantic 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


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) 9-28Semantic 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) 10-28Semantic 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


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) 12-28Semantic Cities @ AAAI 2012

A Big Picture of ICA

A Big Picture

Figure.3: A big picture of ICA (Gargari and Lucas, 2007)


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

Nois there

a colony in an empirewhich has lower cost

than that of imperialist

Exchange the positions of that Imperialist & colony

Initialize the Empires

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


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 15-28Semantic 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 statesby using nominal propagation method

Detect the Congestion AreaCompute Distance between all Aircrafts

in Congestion Area

Define Problem ParametersAnd other Traffic Parameters

16-28

Send new Flight Paths Plan to ExistingAircrafts in Congestion Area

End

Figure.5: Block diagram of our proposed model

Semantic Cities @ AAAI 2012Our 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 17-28Semantic Cities @ AAAI 2012

Figure.6: Graphical display of an exampleconflict 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 dimensionConflict Detection 18-28Semantic 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 19-28Semantic 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 20-28Semantic 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 21-28Semantic 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


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


Test ResultsTable 1: Test results of applying the algorithm onto input graphs

(congestion areas) with specified parameters

Test Results 24-28Semantic 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 25-28Semantic Cities @ AAAI 2012

References1. 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 26-28Semantic Cities @ AAAI 2012

References9. 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 27-28Semantic Cities @ AAAI 2012

The End & Thanks

The End & Thanks

Thanks for your attention!

semantic cities @ aaai 2012 july 23rd, 2012

Documents