Data Mining in Structural Dynamic Analysis: A Signal Processing Perspective

175 Pages·2019·8.497 MB·English

by  Yun Lai Zhou, Magd Abdel Wahab, Nuno M. M. Maia, Linya Liu, Elói Figueiredo

Yun Lai Zhou · Magd Abdel Wahab ·
Nuno M. M. Maia · Linya Liu ·
Elói Figueiredo   Editors
Data Mining
in Structural
Dynamic
Analysis
A Signal Processing Perspective
Data Mining in Structural Dynamic Analysis
Yun Lai Zhou Magd Abdel Wahab
(cid:129) (cid:129)
Nuno M. M. Maia Linya Liu
(cid:129) (cid:129)
ó
El i Figueiredo
Editors
Data Mining in Structural
Dynamic Analysis
A Signal Processing Perspective
123
Editors
YunLai Zhou MagdAbdel Wahab
Department ofCivil Faculty of Engineering andArchitecture
andEnvironmental Engineering GhentUniversity, Laboratory Soete
National University ofSingapore Ghent, Belgium
Singapore, Singapore
LinyaLiu
Nuno M.M.Maia Engineering Research Centerof Railway
Department ofMechanical Engineering EnvironmentVibration andNoise, Ministry
Technical University of Lisbon of Education
Lisbon, Portugal EastChinaJiaotong University
Nanchang, China
Elói Figueiredo
Faculty of Engineering
LusófonaUniversity
Lisbon, Portugal
ISBN978-981-15-0500-3 ISBN978-981-15-0501-0 (eBook)
https://doi.org/10.1007/978-981-15-0501-0
©SpringerNatureSingaporePteLtd.2019
Thisworkissubjecttocopyright.AllrightsarereservedbythePublisher,whetherthewholeorpart
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
orinformationstorageandretrieval,electronicadaptation,computersoftware,orbysimilarordissimilar
methodologynowknownorhereafterdeveloped.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publicationdoesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfrom
therelevantprotectivelawsandregulationsandthereforefreeforgeneraluse.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, expressed or implied, with respect to the material contained
hereinorforanyerrorsoromissionsthatmayhavebeenmade.Thepublisherremainsneutralwithregard
tojurisdictionalclaimsinpublishedmapsandinstitutionalaffiliations.
ThisSpringerimprintispublishedbytheregisteredcompanySpringerNatureSingaporePteLtd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore
Preface
Thestructuralhealthmonitoring(SHM)fieldisconcernedwithincreasingdemand
forimprovedandmorecontinuousconditionassessmentofinfrastructurestobetter
face the challenges presented by modern societies. Thus, the applicability of
computer science techniques for SHM applications has attracted the attention of
researchers and practitioners in the last two decades.
ThisbookaimstodefineandsummarizetheapplicationofdataminingforSHM
of infrastructures, including bridges, railway transport systems, wind turbines,
buildings, and so on. Data mining comprises selecting the appropriate models and
applyingthoseforsearchingpatternsinthefeaturesextractedfrommonitoringdata.
Herein, the machine learning algorithms play an important role to unveil hidden
patterns. The output of the trained algorithms is used to infer knowledge and to
support the decision-making process.
Data mining has been a research hotspot in computer science, in the past dec-
ades, and it demonstrates a bright future in other fields like civil and mechanical
engineering, as it focuses on finding useful information in the data, explaining the
theories hidden behind engineering phenomena, which can be used to generate
knowledge about the structural state condition.
Various engineers from different fields, who are specialized in the data mining,
wereinvitedtowriteallthechapters.Therefore,thisbookisexpectedtoprovidea
common ground for beginners in the fields of structural health monitoring and
structural dynamic analysis.
Singapore, Singapore Yun Lai Zhou
Ghent, Belgium Magd Abdel Wahab
Lisbon, Portugal Nuno M. M. Maia
Nanchang, China Linya Liu
Lisbon, Portugal Elói Figueiredo
v
Contents
Damage Detection for Structural Health Monitoring of Bridges
as a Knowledge Discovery in Databases Process . . . . . . . . . . . . . . . . . . 1
Moisés Silva, Adam Santos and Elói Figueiredo
Structural Health Monitoring of Periodic Infrastructure:
A Review and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Junfang Wang and Jian-Fu Lin
Railway Wheel Out-of-Roundness and Its Effects on Vehicle–Track
Dynamics: A Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Xiao-Zhou Liu
Dynamic Response Analysis of Wind Turbines Under Long-Period
Ground Motions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Wanrun Li, Jie Huang and Yongfeng Du
IntelligentImageAnalysisTechnologyand Applicationfor RailTrack
Inspection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Peng Dai, Shengchun Wang and Zichen Gu
Investigation on the Effect of High-Frequency Torsional Impacts
on the Torsional Vibration of an Oilwell Drill String in Slip Phase . . . . 101
Liping Tang, Xiaohua Zhu and Yunlai Zhou
Carriage–Rail–Viaduct Coupling Analysis Using Dynamic
Flexibility Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Linya Liu, Zhiyuan Zuo, Yun Lai Zhou and Jialiang Qin
Dynamic Response Reconstruction to Supplement the Data
Insufficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
Chaodong Zhang, Jia He and Xiaohua Zhang
A Proposed Method for the Use of the IBIS-FS in Experimental
Modal Analysis of Buildings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Massoud Sofi, Elisa Lumantarna, Priyan Mendis and Lihai Zhang
vii
About the Editors
Yun Lai Zhou Universidade Lusofona, Portugal.
YunLaiZhouisAdjunctProfessorinUniversidadeLusofona,Portugal,andalso
serves as Research Fellow in the Department of Civil and Environmental
Engineering at Hong Kong Polytechnic University, Hong Kong SAR, China. He
obtained his Ph.D. in 2015, in civil engineering from Technical University of
Madrid, Spain. He has authored and co-authored more than 50 scientific publica-
tions in international journals and conference proceedings on the subject of struc-
turalhealthmonitoringandstructuraldynamicanalysis,fracturemechanics,healso
co-edited three books in structural health monitoring and data mining in structural
dynamicanalysis,andhisresearchinterestsincludestructuraldynamicanalysisand
vibration mitigation, structural health monitoring, system identification, and frac-
ture mechanics. e-mail: [email protected]
Magd Abdel Wahab Department of Electrical Energy, Metals, Mechanical
Constructions and Systems, Ghent University, Belgium.
MagdAbdelWahabisProfessorofappliedmechanicsintheSoeteLaboratoryat
Ghent University, Belgium. He received his B.Sc., 1988, in civil engineering and
his M.Sc., 1991, in structural mechanics, both from Cairo University, Egypt. He
completedhisPh.D.infracturemechanicsin1995fromKULeuven,Belgium.He
was awarded the degree of Doctor of Science from the University of Surrey in
2008. He has published more than 210 scientific papers and technical reports in
solid mechanics and dynamics of structures. His research interests include fracture
mechanics, damage mechanics, fatigue of materials, durability, and dynamics and
vibration of structures. e-mail: [email protected]
Nuno M. M. Maia Instituto Superior Tecnico, University of Lisbon, Portugal.
Nuno M. M. Maia had his habilitation in mechanical engineering in 2001 from
Instituto Superior Tecnico, University of Lisbon, Portugal. He obtained his first
degree in 1978 and master's degree in 1985, both in mechanical engineering from
Instituto Superior Tecnico, University of Lisbon. He received his Ph.D. in
mechanicalvibrations(1989)fromImperialCollegeLondon,UK.Hehasauthored
ix
x AbouttheEditors
andco-authoredtwotextbooksandabout200scientificpublicationsininternational
journalsandconferenceproceedingsonthesubjectofmodalanalysisandstructural
dynamics.HeisAssociateEditoroftheShockandVibrationJournal,belongstothe
EditorialAdvisoryBoardoftheJournalofVibrationandControl,isMemberofthe
Editorial Board of the Mechanical Systems and Signal Processing, Member of the
Society for Experimental Mechanics (SEM), Member of the International Institute
of Acoustics and Vibration (IIAV), and Member of the Portuguese Society of
Acoustics (SPA), where he is responsible for the area of vibrations. He has par-
ticipated in and coordinated various national and international research projects in
theareaofmodalanalysisandstructuralvibrations,andhasbeenresponsibleforthe
organization of the International Conference on Structural Engineering Dynamics
(ICEDyn), since 2002. His current research interests are modal analysis and
modal testing, updating of finite element models, coupling and structural modifi-
cation, damage detection in structures, modeling of damping, transmissibility in
multipledegree-of-freedomsystems,andforceidentification.e-mail:nuno.manuel.
[email protected]
Linya Liu Engineering Research Center of Railway Environment Vibration and
Noise, Ministry of Education, East China Jiaotong University, China.
Linya Liu is Professor of rail system noise and vibration control in East China
Jiaotong University, China. He received his B.Sc., 1996, in Southwest Jiaotong
University, China; he completed his Ph.D. in2006 from Tongji University, China,
both in highway and railway engineering. He dedicates his research in the field of
static and dynamic structural analysis, vibration mitigation, and noise control,
especially for the rail system components including floating track slab. He has
authoredandco-authoredseveralbooksandmorethan60scientificpublicationsin
international journals and conference proceedings. e-mail: [email protected]
Elói Figueiredo Faculty of Engineering, Lusófona University, Portugal.
Elói Figueiredo received his Ph.D. in civil engineering in 2010, M.Sc. in
structures of civil engineering in 2007, and licentiate in civil engineering in 2004.
Hestartedteachingin2010attheCatholicUniversityofPortugal,Portugal,andis
currentlyAssociateProfessorintheFacultyofEngineeringofLusófonaUniversity
inLisbon, Portugal. HeisalsoTechnicalConsultantinthe fieldofmonitoring and
maintenance of bridges. He has dedicated his academic career teaching courses in
the field of static and dynamic structural analysis, soil mechanics, foundation
engineering design, seismic engineering, as well as design of reinforced and pre-
stressed concrete structures. In terms of scientific research, he has mainly worked
on structural health monitoring and maintenance of bridges, especially on
vibration-based structural identification and evaluation. He has published over
70 scientific publications in international journals and conference proceedings.
e-mail: eloi.fi[email protected]
Damage Detection for Structural Health
Monitoring of Bridges as a Knowledge
Discovery in Databases Process
MoisésSilva,AdamSantosandElóiFigueiredo
Abstract The structural health monitoring (SHM) field is concerned with the
increasingdemandforimprovedandmorecontinuousconditionassessmentofengi-
neeringinfrastructurestobetterfacethechallengespresentedbymodernsocieties.
Thus, the applicability of computer science techniques for SHM applications has
attractedtheattentionofresearchersandpractitionersinthelastfewyears,especially
to detect damage in structures under operational and environmental conditions. In
the SHM for bridges, the damage detection can be seen as the end of a process to
extract knowledge regarding the structural state condition from vibration response
measurements. In that sense, the damage detection has some similarities with the
KnowledgeDiscoveryinDatabases(KDD)process.Therefore,thischapterintends
toposedamagedetectioninbridgesinthecontextoftheKDDprocess,wheredata
transformationanddataminingplaymajorroles.TheapplicabilityoftheKDDfor
damagedetectionisevaluatedonthewell-knownmonitoringdatasetsfromtheZ-24
Bridge,whereseveraldamagescenarioswerecarriedoutundersevereoperational
andenvironmentaleffects.
· ·
Keywords Knowledgediscoveryindatabasesprocess Datamining Damage
· · ·
identification Bridgemonitoring Vibrationmeasurements Structuralhealth
monitoring
M.Silva
AppliedElectromagnetismLaboratory,UniversidadeFederaldoPará,R.AugustoCorrêa,
Guamá01,Belém,Pará66075-110,Brazil
A.Santos
FacultyofComputingandElectricalEngineering,UniversidadeFederaldoSuleSudeste
doPará,F.17,Q.4,L.E.,Marabá,Pará68505-080,Brazil
B
E.Figueiredo( )
FacultyofEngineering,UniversidadeLusófonadeHumanidadeseTecnologias,
CampoGrande376,1749-024Lisbon,Portugal
e-mail:eloi.fi[email protected]
CONSTRUCT—InstituteofR&DinStructuresandConstruction,R.Dr.RobertoFriass/n,
4200-465Porto,Portugal
©SpringerNatureSingaporePteLtd.2019 1
Y.L.Zhouetal.(eds.),DataMininginStructuralDynamicAnalysis,
https://doi.org/10.1007/978-981-15-0501-0_1
2 M.Silvaetal.
1 Introduction
Thebridgesplayacrucialroleinmodernsocieties,regardlessofculture,geograph-
icallocation,oreconomicdevelopment.Thesafest,economical,andmostdurable
bridges are those that are well managed and maintained. Health monitoring repre-
sentsanimportanttoolinmanagementactivitiesasitpermitsonetoidentifyearly
and progressive structural damage [15]. Overall, the massive data obtained from
monitoringsystemsmustbetransformedtomeaningfulinformationtosupportplan-
ninganddesigningofmaintenanceactivities,increasethesafety,verifyhypotheses,
reduceuncertainty,andwidentheknowledgeandinsightconcerningthemonitored
structure.
Therefore, the process of implementing an autonomous damage identification
strategy for civil, mechanical, and aerospace engineering infrastructure is tradi-
tionally referred to as structural health monitoring (SHM) [17]. The SHM pro-
cess involves the observation of a structure over time using periodically sampled
responsemeasurementsfromanarrayofsensors,theextractionofdamage-sensitive
features from these measurements, and the statistical analysis of these features to
infertheactualstructuralcondition.Thestructuralconditionevaluationconceptgen-
erallyfollowsthedamageidentificationhierarchycomposedoffivelevels(Sect.2.1).
Acknowledgingtheexistenceofotherlevelsandtheirimportanceforamorecompre-
hensiveinformationonthepresenceofdamage,thischapterposestheSHMprocess
mostly in the context of the first level—damage detection. It is important to note
that,nowadays,theSHMiscertainlyoneofthemostpowerfultoolsforinfrastruc-
tureassetmanagement,asitcanprovideinformationfromthestructuralcondition
inrealtime,whichimprovesthedecisionmakingregardingmaintenanceactivities.
Thedamagedetectioncanalsobeseenastheendofaprocesstoextractknowl-
edgeregardingthestructuralstateconditionfromvibrationresponsemeasurements
(e.g., displacements and accelerations). Therefore, the entire process for damage
detectionhassomesimilaritieswiththeKnowledgeDiscoveryinDatabases(KDD)
process (Sect.2.2). Basically, KDD is the overall process of extracting and reveal-
ingusefulinformationfrommassiverawdatameasurementsandtransformsitinto
knowledgefordecisionmaking.Thisprocessisbroadlyappliedinmanyfields,such
ashealthcareindustry,financialmarket,advertising,bioinformatics,astronomy,and
others[2, 3, 12, 27–29, 35].TheKDDprocesscanbebrokendownintofivesteps
[18]:dataselection,datapreprocessing,datatransformation,datamining,anddata
interpretationandevaluationforknowledgediscovery.Eventhoughallthosesteps
areimportantforknowledgegeneration,thischaptergivesmoreattentiontothedata
transformation (Sect.3) and data mining (Sect.4) steps, as those are more directly
relatedtothedamagedetection.
In most recent years, some concepts of KDD have already been posed in the
context of statistical pattern recognition (SPR) paradigm for SHM [17]. The SPR
paradigmiscomposedoffourstages:operationalevaluation,dataacquisition,feature
extraction,andstatisticalmodelingforfeatureclassification[21].Asacomparison,
thedatatransformationstepisrelatedtothefeatureextractionstage,whereseveral