Diagnosis of Process Nonlinearities and Valve Stiction: Data Driven Approaches (Advances in Industrial Control) PDF
Preview Diagnosis of Process Nonlinearities and Valve Stiction: Data Driven Approaches (Advances in Industrial Control)
Advances in Industrial Control Othertitlespublishedinthisseries: DigitalControllerImplementationand ModellingandControlofMini-Flying Fragility Machines RobertS.H.IstepanianandJamesF. PedroCastillo,RogelioLozanoand Whidborne(Eds.) AlejandroDzul OptimisationofIndustrialProcessesat ShipMotionControl SupervisoryLevel TristanPerez DorisSa´ez,AldoCiprianoandAndrzejW. HardDiskDriveServoSystems(2ndEd.) Ordys BenM.Chen,TongH.Lee,KemaoPeng RobustControlofDieselShipPropulsion andVenkatakrishnanVenkataramanan NikolaosXiros Measurement,Control,andCommunication HydraulicServo-systems UsingIEEE1588 MohieddineJelaliandAndreasKroll JohnC.Eidson Model-basedFaultDiagnosisinDynamic PiezoelectricTransducersforVibration SystemsUsingIdentificationTechniques ControlandDamping SilvioSimani,CesareFantuzziandRonJ. S.O.RezaMoheimaniandAndrewJ. Patton Fleming StrategiesforFeedbackLinearisation ManufacturingSystemsControlDesign FreddyGarces,VictorM.Becerra, StjepanBogdan,FrankL.Lewis,Zdenko ChandrasekharKambhampatiandKevin KovaiandJosMirelesJr. Warwick WindupinControl RobustAutonomousGuidance PeterHippe AlbertoIsidori,LorenzoMarconiand AndreaSerrani NonlinearH2/H8ConstrainedFeedback Control DynamicModellingofGasTurbines MuradAbu-Khalaf,JieHuangandFrankL. GennadyG.KulikovandHaydnA. Lewis Thompson(Eds.) ControlofFuelCellPowerSystems PracticalGrey-boxProcessIdentification JayT.Pukrushpan,AnnaG.Stefanopoulou TorstenBohlin andHueiPeng ControlofTrafficSystemsinBuildings FuzzyLogic,IdentificationandPredictive SandorMarkon,HajimeKita,HiroshiKise Control andThomasBartz-Beielstein JairoEspinosa,JoosVandewalleand WindTurbineControlSystems VincentWertz FernandoD.Bianchi,HernnDeBattistaand OptimalReal-timeControlofSewer RicardoJ.Mantz Networks AdvancedFuzzyLogicTechnologiesin MagdaleneMarinakiandMarkos IndustrialApplications Papageorgiou YingBai,HanqiZhuangandDaliWang ProcessModellingforControl (Eds.) BenoˆıtCodrons PracticalPIDControl ComputationalIntelligenceinTimeSeries AntonioVisioli Forecasting AjoyK.PalitandDobrivojePopovic (continuedafterIndex) M. A. A. Shoukat Choudhury • Sirish L. Shah • Nina F. Thornhill Diagnosis of Process Nonlinearities and Valve Stiction Data Driven Approaches 123 M.A.A.ShoukatChoudhury SirishL.Shah BangladeshUniversityof UniversityofAlberta EngineeringandTechnology(BUET) DepartmentChemical&Materials DepartmentofChemicalEngineering Engineering Dhaka-1000 EdmontonABT6G2G6 Bangladesh Canada [email protected] NinaF.Thornhill ImperialCollegeLondon DepartmentofChemicalEngineering SouthKensingtonCampus LondonSW72AZ ISBN:978-3-540-79223-9 e-ISBN:978-3-540-79224-6 LibraryofCongressControlNumber:2008928126 (cid:2)c 2008Springer-VerlagBerlinHeidelberg Thisworkissubjecttocopyright.Allrightsarereserved,whetherthewholeorpartofthematerialis concerned,specificallytherightsoftranslation,reprinting,reuseofillustrations,recitation,broadcasting, reproductiononmicrofilmorinanyotherway,andstorageindatabanks.Duplicationofthispublication orpartsthereofispermittedonlyundertheprovisionsoftheGermanCopyrightLawofSeptember9, 1965,initscurrentversion,andpermissionforusemustalwaysbeobtainedfromSpringer.Violationsare liabletoprosecutionundertheGermanCopyrightLaw. Theuseofgeneraldescriptivenames,registerednames,trademarks,etc.inthispublicationdoesnotimply, evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfromtherelevantprotectivelaws andregulationsandthereforefreeforgeneraluse. Coverdesign:eStudioCalamarS.L. Printedonacid-freepaper 9 8 7 6 5 4 3 2 1 springer.com Toallourmentorsandstudents Series Editors’ Foreword The series Advances in Industrial Control aims to report and encourage technology transfer in control engineering. The rapid development of control technology has an impact on all areas of the control discipline. New theory, new controllers, actuators, sensors, new industrial processes, computer methods, new applications, new philosophies…, new challenges. Much of this development work resides in industrial reports, feasibility study papers and the reports of advanced collaborative projects. The series offers an opportunity for researchers to present an extended exposition of such new work in all aspects of industrial control for wider and rapid dissemination. Statistical process control was probably the first complete process-monitoring paradigm. It was conceived to detect when process plant conditions had changed and process outputs were no longer meeting the desired product specifications. The underlying science of statistical process control was based on Gaussian statistical analysis and principles and the method also involves tools designed to uncover or diagnose the source of the detected process change. These diagnostic tools include methods such as “cause and effect” analysis and the associated fishbone charts; however, diagnosing the cause of process change was found to be a much more difficult problem than simply identifying that process change had occurred. In the context of closed-loop control system assessment, the seminal work of Harris and co-workers (1989) initiated new practical methods of determining whether process control loops are well tuned. A key motivation for this work was to answer the question often raised by industrial engineers of how to find out (efficiently and economically) whether any of the thousands of PID control loops present in a typical industrial plant are poorly tuned and need to be retuned. To achieve this, the solutions proposed exploited the extensive on-line data that process computers now store routinely for future analysis. Process control equipment and software vendors enthusiastically took up the approaches developed and quite a few controller assessment products are now available commercially. Academic researchers also found performance assessment to be a fruitful area of investigation and the Advances in Industrial Control monograph series published the work of Biao Huang and Sirish Shah as Performance Assessment of Control Loops (ISBN 978-1-85233-639-4, 1999). More recent contributions to this field vii viii Series Editors’ Foreword were published in the series as the contributed volume, Process Control Performance Assessment: From Theory to Implementation with Andrzej Ordys, Damian Uduehi, and Michael Johnson as Editors (ISBN 978-1-84628-623-0, 2007). Along with this good progress in process controller assessment methods, researchers have also been investigating techniques to diagnose what is causing the process or control loop degradation. This requires the use of on-line data to identify faults via new diagnostic indicators of typical process problems. A significant focus of some of this research has been the issue of valve problems; a research direction that has been motivated by some industrial statistics that show up to 40% of control loops having performance degradation attributable to valve problems. Shoukat Choudhury, Sirish Shah, and Nina Thornhill have been very active in this research field for a number of years and have written a coherent and consistent presentation of their many research results as this monograph, Diagnosis of Process Nonlinearities and Valve Stiction. The Advances in Industrial Control series is pleased to welcome this new and substantial contribution to the process diagnostic literature. The reader will find the exploitation of the extensive process data archives created by today’s process computer systems one theme in the monograph. From another viewpoint, the use of higher-order statistics could be considered to provide a continuing link to the earlier methods of the statistical process control paradigm. Then there is the significant work presented on control valves, stiction, and plant-wide disturbances and this demonstrates important new diagnostic tools that can be used by the industrial process engineer. Finally, the volume has archival value since it brings together so many new practical results in one consistent framework. The result is a stimulating volume in the Advances in Industrial Control series that will be appreciated and used by industrial process control engineers, and explored further by postgraduate students and control experts from the academic controls community. Glasgow M.J. Grimble Scotland, UK M.A. Johnson Preface Classicalcontrolhasmaturedtothepointwhereoff-the-shelfcontrolsolutions,for linear time-invariant systems, are easily available from most major control ven- dors. In the relentless drive for improved regulatory performance due to increas- ing energy costs, higher quality specifications and competition, advanced tools for monitoring closed-loop performance are becoming available. Such measures of closed-loop control performance can now be readily computed in an on-line manner. Performancemetrics,fromroutineoperatingdata,giveanindicationofclosed- loopperformancewithrespecttouser-definedbenchmarks,suchasminimumvari- ancebenchmark,orahistoricalbenchmarkofsatisfactoryperformanceoradesired settlingtimebenchmark.Conversely,suchbenchmarksalsoprovideanindicationof thedifficultyofcontrollingaprocess.However,thediagnosisofpoorperformance isanareawithmanychallenges.Poorclosed-systemperformancemaybeduetoa varietyofreasons,forexample,poorcontrollertuning,impropercontrollerconfigu- ration,processoperatingconditionchanges,processconstraints,actuatorproblems suchasvalvestictionorsaturationandsoon.Thediagnosisofpoorperformanceis anon-trivialproblem. Access to routine process data is now a norm. The purpose of this book is to introducenoveltoolsintheanalysisofsuchdatatodeterminethediagnosisofpoor controllerperformance.Themainmaterialinthebookisconcernedwiththedetec- tion, diagnosis and quantification of process nonlinearities. Process changes such as saturation constraints, sticky valves and nonlinear operating regimes introduce nonlinearitiesincontrolsignals.Theveryspecificfocusofthisbookisondetecting processnonlinearitiesanddiagnosingandquantificationofthesourceofsuchnon- linearities. The main tool for the detection of such nonlinearities are higher-order statistics, in particular third-order cumulants and the bispectrum, as the spectral counterpartofthethird-ordermomentsofthedistributionofasignal.Specifically, thistextprovidesanswerstoquestionssuchasthefollowing: • Are the process data well-conditioned for analysis in the first place? Well- conditioned data are an indispensable part of computing any performance met- ric.Itisfutiletocomputeperformancemetricsandmakediagnosisconclusions on poorly conditioned data, as the results may lead to erroneous conclusions. Thisbookdwellsoncheckingtheintegrityofarchiveddatapriortousingsuch ix x Preface data for computing performance metrics. For example, methods for estimating datacompressionfactorsandsensorquantizationestimatesareconsideredinthis book. • Howtodetermineameasureofprocessnonlinearityfromprocessdata?Sucha metric may indeed give a clue to the degree of difficulty in controlling a given process.Iftheprocessisdiagnosedtobestronglynonlinearthanthesuggested remedialactionwouldbetoconsideranonlinearcontrolalgorithm. • Howtocheckiftheactuatorisperformingsatisfactorily?Statisticsfromseveral large-scaleindustrialstudieshaveconfirmedthatasmanyas40%ofthecontrol loopsdonotperformwellbecauseofvalveproblems.Inthisrespect,animpor- tantpartofthisbookisfocussedondevelopingareliabletestforthedetectionof valvestiction.Asignificantportionofthismonographisdevotedtowardsdata- based modelling of valve dynamics and the detection and diagnosis of valve or actuator problems from routine operating data. Many simulation examples and over a dozen industrial case studies are provided to illustrate the application of higher-orderstatisticaltoolsdevelopedinthisbook. • Are there plant-wide oscillations and if so how to detect loops with common oscillationsanddiagnosethepossiblesourceoftheoscillations?Itisnotuncom- mon,asevidentfromthethreeindustrialcasestudiesconsideredhere,formany loops to have common oscillations. Because of the highly integrated nature of modern plants, oscillations that may arise in one loop due to overly tuned con- trollergainsorastickyvalveoranoscillatorydisturbance,caneasilypropagate tomanyotherloopsdownstreamaswellasupstreamduetothepresenceofre- cycles.Newlydevelopedtoolsforthedetectionanddiagnosisofsuchproblems arepresentedtowardstheendofthisbook. Mostoftheresultspresentedherehaveappearedinthearchivalliterature,albeit in compact forms, over the last several years. One motivation in preparing this monograph was to consolidate all these results in a logical and presentable form in one place. In this respect, the book is likely to be of use for new researchers asaplacetoaccessbasicinformationpresentedinatutorialformatasspaceisno longerarestriction,asitisintechnicalpapersinarchivedjournalsandconference proceedings. However, there are results here that will also appeal to the practising engineer.Portionsofthisbookcanalsobeusedinagraduate-levelcourseinprocess control,signalprocessingandfaultdetectionandisolation.Manyofthechaptersdo notrequireknowledgeofearlierchaptersasprerequisiteinformationandtherefore canbereadindependently. Theworkinthisbookhasbeentheoutcomeofseveralyearsofresearcheffortsby theauthorswithcontributionsfrommanyothergraduatestudentsattheUniversity ofAlberta,Canada,andintheCentreforProcessSystemsEngineeringatUniversity College London and Imperial College London, UK, as well as collaborative work with industry. It is a pleasure to thank the many people who have contributed so generouslytotheconceptionandcreationofthematerialinthisbook.Theresearch environment at U of A, UCL and Imperial College has been very stimulating and superblyconducivetowardsdoingcollaborativeandcross-disciplinaryresearchand has provided much needed fuel to the efforts that underwent in the work reported
Description:The list of books you might like
