Computational intelligence in integrated airline scheduling

210 Pages·2009·19.8 MB·English

by  Tobias Grosche

ValentinaEmiliaBalas,Ja´nosFodor,andAnnama´riaR.Va´rkonyi-Ko´czy(Eds.)
SoftComputingBasedModelinginIntelligentSystems
StudiesinComputationalIntelligence,Volume196
Editor-in-Chief
Prof.JanuszKacprzyk
SystemsResearchInstitute
PolishAcademyofSciences
ul.Newelska6
01-447Warsaw
Poland
E-mail:[email protected]
Furthervolumesofthisseriescanbefoundonourhomepage: Vol.185.AnthonyBrabazonandMichaelO’Neill(Eds.)
springer.com NaturalComputinginComputationalFinance,2009
ISBN978-3-540-95973-1
Vol.173.TobiasGrosche Vol.186.Chi-KeongGohandKayChenTan
ComputationalIntelligenceinIntegratedAirlineScheduling, EvolutionaryMulti-objectiveOptimizationinUncertain
2009 Environments,2009
ISBN978-3-540-89886-3 ISBN978-3-540-95975-5
Vol.174.AjithAbraham,RafaelFalco´nandRafaelBello(Eds.) Vol.187.MitsuoGen,DavidGreen,OsamuKatai,BobMcKay,
RoughSetTheory:ATrueLandmarkinDataAnalysis,2009 AkiraNamatame,RuhulA.SarkerandByoung-TakZhang
ISBN978-3-540-89886-3 (Eds.)
IntelligentandEvolutionarySystems,2009
Vol.175.GodfreyC.OnwuboluandDonaldDavendra(Eds.)
ISBN978-3-540-95977-9
DifferentialEvolution:AHandbookforGlobal
Permutation-BasedCombinatorialOptimization,2009 Vol.188.AgustínGutiérrezandSantiagoMarco(Eds.)
ISBN978-3-540-92150-9 BiologicallyInspiredSignalProcessingforChemicalSensing,
2009
Vol.176.BeniaminoMurgante,GiuseppeBorrusoand
ISBN978-3-642-00175-8
AlessandraLapucci(Eds.)
GeocomputationandUrbanPlanning,2009 Vol.189.SallyMcClean,PeterMillard,EliaEl-Darziand
ISBN978-3-540-89929-7 ChrisNugent(Eds.)
IntelligentPatientManagement,2009
Vol.177.DikaiLiu,LingfengWangandKayChenTan(Eds.) ISBN978-3-642-00178-9
DesignandControlofIntelligentRoboticSystems,2009
ISBN978-3-540-89932-7 Vol.190.K.R.Venugopal,K.G.SrinivasaandL.M.Patnaik
SoftComputingforDataMiningApplications,2009
Vol.178.SwagatamDas,AjithAbrahamandAmitKonar ISBN978-3-642-00192-5
MetaheuristicClustering,2009
ISBN978-3-540-92172-1 Vol.191.ZongWooGeem(Ed.)
Music-InspiredHarmonySearchAlgorithm,2009
Vol.179.MirceaGh.NegoitaandSorinHintea ISBN978-3-642-00184-0
Bio-InspiredTechnologiesfortheHardwareofAdaptiveSystems,
2009 Vol.192.AgusBudiyono,BambangRiyantoandEndraJoelianto
ISBN978-3-540-76994-1 (Eds.)
IntelligentUnmannedSystems:TheoryandApplications,2009
Vol.180.WojciechMitkowskiandJanuszKacprzyk(Eds.) ISBN978-3-642-00263-2
ModellingDynamicsinProcessesandSystems,2009
ISBN978-3-540-92202-5 Vol.193.RaymondChiong(Ed.)
Nature-InspiredAlgorithmsforOptimisation,2009
Vol.181.GeorgiosMiaoulisandDimitriPlemenos(Eds.) ISBN978-3-642-00266-3
IntelligentSceneModellingInformationSystems,2009
ISBN978-3-540-92901-7 Vol.194.IanDempsey,MichaelO’NeillandAnthonyBrabazon
(Eds.)
Vol.182.AndrzejBargielaandWitoldPedrycz(Eds.) FoundationsinGrammaticalEvolutionforDynamic
Human-CentricInformationProcessingThroughGranular Environments,2009
Modelling,2009 ISBN978-3-642-00313-4
ISBN978-3-540-92915-4
Vol.195.VivekBannoreandLeszekSwierkowski
Vol.183.MarcoA.C.PachecoandMarleyM.B.R.Vellasco(Eds.) Iterative-InterpolationSuper-ResolutionImageReconstruction:
IntelligentSystemsinOilFieldDevelopmentunderUncertainty, AComputationallyEfficientTechnique,2009
2009 ISBN978-3-642-00384-4
ISBN978-3-540-92999-4
Vol.196.ValentinaEmiliaBalas,Ja´nosFodorand
Vol.184.LjupcoKocarev,ZbigniewGaliasandShiguoLian Annama´riaR.Va´rkonyi-Ko´czy(Eds.)
(Eds.) SoftComputingBasedModeling
IntelligentComputingBasedonChaos,2009 inIntelligentSystems,2009
ISBN978-3-540-95971-7 ISBN978-3-642-00447-6
Valentina Emilia Balas
Ja´nos Fodor
Annama´ria R.Va´rkonyi-Ko´czy (Eds.)
Soft Computing Based Modeling
in Intelligent Systems
123
Prof.ValentinaE.Balas Prof.Annama´riaR.Va´rkonyi-Ko´czy
DepartmentofAutomationandApplied DepartmentofMeasurementandInformation
Informatics, Systems
FacultyofEngineering, BudapestUniversityofTechnologyand
“AurelVlaicu"UniversityfromArad Economics
B-dulRevolutiei77 MagyarTudosokkrt.2.
310130Arad H-1521Budapest
Romania Hungary
E-mail:[email protected] E-mail:[email protected]
Prof.Ja´nosFodor
InstituteofIntelligentEngineeringSystems
JohnvonNeumannFacultyofInformatics
BudapestTech
Be´csiu´t96/b
H-1034Budapest
Hungary
E-mail:[email protected]
ISBN 978-3-642-00447-6 e-ISBN978-3-642-00448-3
DOI 10.1007/978-3-642-00448-3
Studiesin Computational Intelligence ISSN1860949X
Library of Congress Control Number:2009921161
(cid:1)c 2009 Springer-Verlag Berlin Heidelberg
This work is subject to copyright. All rights are reserved, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilm or in any other way,
andstorageindatabanks.Duplicationofthispublicationorpartsthereofispermitted
only under the provisions of the German Copyright Law of September 9, 1965, in
its current version, and permission for use must always be obtained from Springer.
Violations are liable to prosecution undertheGerman Copyright Law.
Theuseofgeneral descriptivenames,registered names,trademarks,etc.inthispubli-
cationdoesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesare
exemptfromtherelevantprotectivelawsandregulationsandthereforefreeforgeneral
use.
Typeset&CoverDesign:ScientificPublishing Services Pvt. Ltd., Chennai, India.
Printed in acid-free paper
9 8 7 6 5 4 3 2 1
springer.com
Preface
The book “Soft Computing Based Modeling in Intelligent Systems”contains the ex-
tended works originally presented at the IEEE International Workshop SOFA 2005
and additional papers.
SOFA, an acronym for SOFt computing and Applications, is an international work-
shop intended to advance the theory and applications of intelligent systems and soft
computing.
Lotfi Zadeh, the inventor of fuzzy logic, has suggested the term “Soft Computing.”
He created the Berkeley Initiative of Soft Computing (BISC) to connect researchers
working in these new areas of AI. Professor Zadeh participated actively in our work-
shop.
Soft Computing techniques are tolerant to imprecision, uncertainty and partial truth.
Due to the large variety and complexity of the domain, the constituting methods of
Soft Computing are not competing for a comprehensive ultimate solution. Instead they
are complementing each other, for dedicated solutions adapted to each specific prob-
lem. Hundreds of concrete applications are already available in many domains. Model
based approaches offer a very challenging way to integrate a priori knowledge into
procedures. Due to their flexibility, robustness, and easy interpretability, the soft com-
puting applications will continue to have an exceptional role in our technologies. The
applications of Soft Computing techniques in emerging research areas show its matur-
ity and usefulness.
The IEEE International Workshop SOFA 2005 held Szeged-Hungary and Arad-
Romania in 2005 has led to the publication of these two edited volumes. This volume
contains Soft Computing methods and applications in modeling, optimisation and
prediction.
Chapter 1 by Nikola Kasabov presents a comparative study of major modeling
and pattern discovery approaches applicable to the area of Bioinformatics and the
area of decision support systems in general. These approaches include inductive
versus transductive reasoning, global, local, and personalized modeling and their
potentials are illustrated on a case study of gene expression and clinical data related
to cancer outcome prognosis. While inductive modeling is used to develop a model
(function) from data on the whole problem space and then to recall it on new data,
transductive modeling is concerned with the creation of single model for every new
input vector based on some closest vectors from the existing problem space. The
VI Preface
paper uses several techniques to illustrate these approaches – multiple linear regres-
sion, Bayesian inference, support vector machines, evolving connectionist systems
(ECOS), weighted kNN – each of them providing different accuracy on specific prob-
lem and facilitating the discovery of different patterns and rules from data.
Chapter 2 by Bernard de Baets introduces an n-ary aggregation useful to determine
a collective decision, preference, or opinion, based on several individual decisions,
preferences, or opinions.
In Chapter 3 by Antonio Ruano the design of inside air temperature predictive neu-
ral network models, to be used for predictive thermal comfort control, is discussed.
The design is based on the joint use of multi-objective genetic (MOGA) algorithms,
for selecting the network structure and the network inputs, and a derivative algorithm,
for parameter estimation. Climate and environmental data from a secondary school
located in the south of Portugal, collected by a remote data acquisition system, are
used to generate the models. By using a sliding window adaptive methodology, the
good results obtained off-line are extended throughout the whole year.
Chapter 4 by Annamaria Varkoniy-Koczy presents some practical solutions of
engineering problems which involve model-integrated computing. Model based ap-
proaches offer a very challenging way to integrate a priori knowledge into the proce-
dure. Due to their flexibility, robustness, and easy interpretability, the application of
soft computing, in particular fuzzy and neural network based models, may have an
exceptional role at many fields, especially in cases where the problem to be solved is
highly nonlinear or when only partial, uncertain and/or inaccurate data is available.
Nevertheless, ever so advantageous their usage can be, it is still limited by their expo-
nentially increasing computational complexity. Although, a possible solution can be, if
we combine soft computing and anytime techniques, because the anytime mode of
operation is able to adaptively cope with the available, usually imperfect or even miss-
ing information, the dynamically changing, possibly insufficient amount of resources
and reaction time.
In this chapter the applicability of (Higher Order) Singular Value Decomposition
based anytime Soft Computational models is analyzed in dynamically changing, com-
plex, time-critical systems.
Chapter 5 by Nathalie Pessel, Jean Duplaix, Jean-François Balmat, Frédéric Lafont
presents a modeling methodology of the complex systems. The system class consid-
ered put together multivariable, non-linear, non-stationary and strongly disturbed
systems. The authors proposed to modelise these systems with multi-model at variable
structure. The methodology is based on a supervision of a multi-model structure de-
fined by a statistical analysis and validated by the expert knowledge of the system.
The association of the expert knowledge and the data analysis allows to detect the
correlation between variables and to select the most significant in each group of corre-
lated variable. Neural Networks are efficient for the complex non-linear dynamic
systems modeling. The management of the multi-model system is carried out by su-
pervision with a Hierarchical Fuzzy Logic. The multi-structure modeling methodology
presented refers to an expert knowledge of the system. This knowledge steps in the
data analysis and more exactly in the definition of the operating ranges number and in
the selection of the pertinent variable in a group of correlated variables. The supervi-
sion definition needs to this expert knowledge to describe the rule bases. The several
Preface VII
experiments presented in this chapter show the performances of this multi-structure
modeling methodology.
Chapter 6 by József Dombi give a generalization of the Dombi operator. This gen-
eralization involves most well-known operators. The operator has only two parameters
and the conjunctive and disjunctive operators differ only in the sign of one of the
parameters. The De Morgan identity is also examined. The author shows that the min-
max and the Dombi operator cases fulfill the De Morgan identity with all rational
strong negations. Finally the author shows the isomorphism between the operator and
the multiplicative utility function.
Chapter 7 by Hajime Nobuhara and Barnab_as Bede a multi-channel representa-
tions of max-plus algebra based wavelet transform are proposed. The multi-channel
representations can be constructed from two kinds of channels, that is, an approxima-
tion channel and detailed channels which are correspond to the maximum element of
pixels and the subtraction between the fixed pixel and other pixels in the sampling
window. Therefore, in the case of max-plus algebra based wavelet transform, the au-
thors can use various sampling windows with arbitrary size and shape to define multi-
channel wavelet transform. The max-plus algebra based wavelet transform has several
advantages, and one of them is suitable for edge image compression due to non-linear
operation (max and min), compared with the conventional wavelets. This property
efficiently works on the predictive-frame compression used in the video-coding.
Furthermore, in the video coding, the authors confirmed that the max-plus algebra
based wavelet transform efficiently works.
Chapter 8 by Marius Balas is stressing one of the fundamental sides of the fuzzy
sets: the interpolative one. Since any fuzzy controller can be approximated by a corre-
sponding interpolative one, the linear interpolations can be fully applied in almost any
fuzzy sets application: in the elaboration of the control rules as well as in the imple-
mentations. On the other hand the interpolative implementations are very feasible in
almost any possible technology. That is why the paper is presenting a methodology
that takes advantage of the fuzzy linguistic conception and the interpolative implemen-
tation in the same time. A case study focused on a car following algorithm is illustrat-
ing the fuzzy-interpolative methodology.
Chapter 9 by Sanda Dale and Toma-Leonida Dragomir explains and exemplifies
how the controllers with interpolative blocks can replace fuzzy controllers in control
structures. This is possible because fuzzy controllers belong also to the interpolative-
type controller category, meaning controllers which implements interpolative-type
reasoning. That kind of replacement is not only a formal operation; it is also associ-
ated with further corrections that confer to the structures with interpolative controllers
enough flexibility to obtain better performances. The possibility of performances
improvement on a flexible structure is the main argument. Another argument is the
reduced calculus time, suited for the real-time implementation - it’s about “look-up
table” type solutions and the possibility to obtain simple controllers with robustness
properties. In order to illustrate the above affirmations, two case studies were devel-
oped: an electromechanical ball and beam nonlinear system and a positioning system
with Lyapunov constraints and state limitations.
We especially thank the honorary chair of the International IEEE SOFA 2005 Prof
Lotfi Zadeh, who motivated us throughout the process. Most of the papers in this book
reflect the extended work from this conference.
VIII Preface
We are grateful to all the authors for enthusiastically submitting high quality work
to this publication and to the reviewers.
A particular acknowledgement goes to Prof. J. Kacprzyk (Editor-in-Chief, Springer
Studies in Computational Intelligence Series), which has given to us the opportunity to
act as Guest Editors of this book.
Finally, we express our special thanks to Springer, in particular, Dr. Thomas Ditz-
inger (Springer Engineering In house Editor, Studies in Computational Intelligence
Series) and Ms. Heather King (Editorial Assistant, Springer Verlag, Heidelberg) for
the editorial assistance and excellent collaboration during the development of this
volume.
We hope that the reader will share our excitement and find the volume “Soft Com-
puting Based Modeling in Intelligent Systems” both inspiring and useful.

December 2008 Valentina Emilia Balas
János Fodor
Annamária R. Várkonyi-Kóczy
Contents
1 Soft Computing Methods for Global, Local and
Personalized Modeling and Applications in Bioinformatics
Nikola Kasabov ................................................... 1
2 Orthosymmetry Aspects of Aggregation Functions
Koen C. Maes, Bernard De Baets................................... 19
3 MOGA Design of Neural Network Predictors of Inside
Temperature in Public Buildings
A.E. Ruano, E.M. Crispim, P.M. Fraza˜o ............................ 35
4 Model Based Anytime Soft Computing Approaches in
Engineering Applications
Annam´aria R. V´arkonyi-Ko´czy ..................................... 63
5 A Multi-structure Modeling Methodology
Nathalie Pessel, Jean Duplaix, Jean-Franc¸ois Balmat, Fr´ed´eric Lafont ... 93
6 The Generalized Dombi Operator Family and the
Multiplicative Utility Function
J´ozsef Dombi..................................................... 115
7 Multi-channel Representations of Max-Plus Algebra Based
Wavelet Transform and Their Application to Video Coding
Hajime Nobuhara, Barnab´as Bede................................... 133
8 The Fuzzy-Interpolative Methodology
Marius M. Balas.................................................. 145
9 Interpolative-Type Control Solutions
Sanda Dale, Toma-Leonida Dragomir................................ 169
Author Index................................................... 205