Advances in Industrial Control Yang Shi Chao Shen Henglai Wei Kunwu Zhang Advanced Model Predictive Control for Autonomous Marine Vehicles Advances in Industrial Control SeriesEditor MichaelJ.Grimble,IndustrialControlCentre,UniversityofStrathclyde,Glasgow, UK EditorialBoard GrahamGoodwin,SchoolofElectricalEngineeringandComputing,Universityof Newcastle,Callaghan,NSW,Australia ThomasJ.Harris,DepartmentofChemicalEngineering,Queen’sUniversity, Kingston,ON,Canada TongHengLee ,DepartmentofElectricalandComputerEngineering,National UniversityofSingapore,Singapore,Singapore OmP.Malik,SchulichSchoolofEngineering,UniversityofCalgary,Calgary,AB, Canada Kim-FungMan,CityUniversityHongKong,Kowloon,HongKong GustafOlsson,DepartmentofIndustrialElectricalEngineeringandAutomation, LundInstituteofTechnology,Lund,Sweden AsokRay,DepartmentofMechanicalEngineering,PennsylvaniaStateUniversity, UniversityPark,PA,USA SebastianEngell,LehrstuhlfürSystemdynamikundProzessführung,Technische UniversitätDortmund,Dortmund,Germany IkuoYamamoto,GraduateSchoolofEngineering,UniversityofNagasaki, Nagasaki,Japan AdvancesinIndustrialControlisaseriesofmonographsandcontributedtitlesfocusingon theapplicationsofadvancedandnovelcontrolmethodswithinappliedsettings.Thisseries hasworldwidedistributiontoengineers,researchersandlibraries. The series promotes the exchange of information between academia and industry, to whichendthebooksalldemonstratesometheoreticalaspectofanadvancedornewcontrol method and show how it can be applied either in a pilot plant or in some real industrial situation.Thebooksaredistinguishedbythecombinationofthetypeoftheoryusedandthe typeofapplicationexemplified.Notethat“industrial”herehasaverybroadinterpretation;it applies not merely to the processes employed in industrial plants but to systems such as avionicsandautomotivebrakesanddrivetrain.Thisseriescomplementsthetheoreticaland moremathematicalapproachofCommunicationsandControlEngineering. IndexedbySCOPUSandEngineeringIndex. Proposalsforthisseries,composedofaproposalform(pleaseaskthein-houseeditorbelow), adraftContents,atleasttwosamplechaptersandanauthorcv(withasynopsisofthewhole project,ifpossible)canbesubmittedtoeitherofthe: SeriesEditor ProfessorMichaelJ.Grimble: Department of Electronic and Electrical Engineering, Royal College Building, 204GeorgeStreet,GlasgowG11XW,UnitedKingdom; e-mail:[email protected] orthe In-houseEditor Mr.OliverJackson: SpringerLondon,4CrinanStreet,London,N19XW,UnitedKingdom; e-mail:[email protected] Proposalsarepeer-reviewed. PublishingEthics Researchers should conduct their research from research proposal to publication in linewithbestpracticesandcodesofconductofrelevantprofessionalbodiesand/ornational andinternationalregulatorybodies.Formoredetailsonindividualethicsmatterspleasesee: https://www.springer.com/gp/authors-editors/journal-author/journal-author-helpdesk/ publishing-ethics/14214 · · · Yang Shi Chao Shen Henglai Wei Kunwu Zhang Advanced Model Predictive Control for Autonomous Marine Vehicles YangShi ChaoShen DepartmentofMechanicalEngineering DepartmentofSystemsandComputer UniversityofVictoria Engineering Victoria,BC,Canada CarletonUniversity Ottawa,ON,Canada HenglaiWei DepartmentofMechanicalEngineering KunwuZhang UniversityofVictoria DepartmentofMechanicalEngineering Victoria,BC,Canada UniversityofVictoria Victoria,BC,Canada ISSN 1430-9491 ISSN 2193-1577 (electronic) AdvancesinIndustrialControl ISBN 978-3-031-19353-8 ISBN 978-3-031-19354-5 (eBook) https://doi.org/10.1007/978-3-031-19354-5 MathematicsSubjectClassification:9302,93C10,93D05,93C85,93C15,93C95 MATLABisaregisteredtrademarkofTheMathWorks,Inc. Seehttps://www.mathworks.com/trademarksforalistofadditionaltrademarks. ©TheEditor(s)(ifapplicable)andTheAuthor(s),underexclusivelicensetoSpringerNature SwitzerlandAG2023 Thisworkissubjecttocopyright.AllrightsaresolelyandexclusivelylicensedbythePublisher,whether thewholeorpartofthematerialisconcerned,specificallytherightsoftranslation,reprinting,reuse ofillustrations,recitation,broadcasting,reproductiononmicrofilmsorinanyotherphysicalway,and transmissionorinformationstorageandretrieval,electronicadaptation,computersoftware,orbysimilar ordissimilarmethodologynowknownorhereafterdeveloped. Theuseofgeneraldescriptivenames,registerednames,trademarks,servicemarks,etc.inthispublication doesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfromtherelevant protectivelawsandregulationsandthereforefreeforgeneraluse. Thepublisher,theauthors,andtheeditorsaresafetoassumethattheadviceandinformationinthisbook arebelievedtobetrueandaccurateatthedateofpublication.Neitherthepublishernortheauthorsor theeditorsgiveawarranty,expressedorimplied,withrespecttothematerialcontainedhereinorforany errorsoromissionsthatmayhavebeenmade.Thepublisherremainsneutralwithregardtojurisdictional claimsinpublishedmapsandinstitutionalaffiliations. ThisSpringerimprintispublishedbytheregisteredcompanySpringerNatureSwitzerlandAG Theregisteredcompanyaddressis:Gewerbestrasse11,6330Cham,Switzerland Series Editor’s Foreword Controlengineeringisviewedratherdifferentlybyresearchersandthosethatmust design, tune, implement, and maintain control systems. Researchers often develop algorithmsforverygeneralproblemswithawell-definedmathematicalbasis;engi- neers have more immediate concerns over the limitations of equipment, quality of control, safety, security, and plant downtime. The series Advances in Industrial Control attempts to bridge this divide by encouraging the adoption of advanced controltechniquesasandwhentheyarelikelytobebeneficial. Therapiddevelopmentofnewcontroltheoryandtechnologyhasanimpactonall areasofengineeringapplications.Thismonographserieshasafocusonapplications that often stimulate the development of new control paradigms. This is desirable if the different aspects of the ‘control design’ problem are to be explored with the samededicationthat‘controlsynthesis’problemshavereceived.Theseriesprovides an opportunity for researchers to present new work and raises awareness of the substantialbenefitsthatadvancedcontrolcanprovidewhiledescribingthechallenges thatcanarise. Thisparticularmonographisconcernedwiththeapplicationofmodelpredictive control(MPC)forautonomousmarinevehicles,includingbothsurfaceandunder- watervehicles.Thisisarapidlygrowingareagiventheimplicationsfordefenseand theimportanceofunderwatercables,pipelines,andstructures.Thetextprovidesan overviewofthedifferentmodelingandcontroldesignproblems. Chapter 1 considers an introduction to control of autonomous marine vehicles andtomodelpredictivecontrol,includingtheuseofthe‘recedinghorizon’control strategy. The different modes of operation, including dynamic positioning, path following, trajectory tracking, and the use of cooperative control, are discussed. Chapter 2 covers the modeling of autonomous marine vehicles involving both the kinematicsanddynamicsofasystem. Chapter3coversthedesignofpath-planningandtrackingcontrolsusinganinte- gratedcontrolphilosophy.Themainproblemconsideredisthedesignofanonlinear- MPC-based tracking controller accounting for the dynamics and kinematics of an autonomous underwater vehicle, including a stability analysis for the closed-loop system. The performance is demonstrated using simulation. Chapter 4 considers v vi SeriesEditor’sForeword Lyapunov-based MPC for the dynamic positioning and trajectory-tracking control design. The optimization problem and a solution methodology are described. An exampleillustratesthetypeofresultsthatcanbeachieved. Chapter 5 considers the path-following problem of an autonomous underwater vehicle,inwhichthepathfollowingisthemaintask,andthespeedprofileisconsid- eredasecondarytask.Amulti-objectiveMPCapproachisfollowed,thatisofinterest. Theresultsareillustratedinasimulationthatconsiderstherobustnessofthesolu- tionandthisisasimportantastheperformanceachieved.Theintroductionofocean currentdisturbancesintotheproblemisconsideredinChap.6.Adistributedmodel predictiveformationofthetrackingcontrolisproposed. Chapter7,onrobustdistributedMPCplatooningcontrolforautonomoussurface vehicles,followsonfromresearchonroadvehiclesystems,andinbothcases,energy can be reduced to extend range and safety can be improved. Chapter 8 deals with theimportantpracticalproblemofimplementationofnonlinearMPCalgorithmsfor trackingcontrol.Theproblemformulationandnumericalsolutionoftheoptimiza- tion problem are discussed, and various designs are compared in simulation. The computingtimesarecompared,androbustnessisnotforgotten. Chapter 9 summarizes the main contributions. It also discusses future research directionsandincludesalookatsomefundamentalquestions,bothofwhichshould beusefultoresearchengineersanddevelopers. Theauthorshaveprovidedausefulintroductiontotheapplicationproblemsand suggestedverysuitablecontrolsolutions.Thistextisinaveryimportantapplication areathatishighlylikelytogrowsubstantially.Insomeapplicationproblems,theuse ofadvancedcontrolsmaybeconsideredoptional,butinareasofautonomousunder- watersystemsthemoreadvancedcontrolmethodsmaybeessentialifperformance goalsaretobemet.Forexample,thereistheneedtolimitbatteryloadingandextend therangeofautonomousvehicles,andthissuggestssomeformofoptimizationwill beneeded.Themonographshouldbeusefultobothvehiclecontroldesignersandto studentsinterestedinthetheoreticalmethodsandtoolsthatcanbeapplied. Glasgow,UK MichaelJ.Grimble August2022 Preface The ocean covers about 72% of the earth’s surface, but most of the area has not yet been explored. There is an increasing demand for advanced technologies and equipment to explore and exploit the ocean for broad applications, such as tidal energyharnessing,deepoceanexploration,oceanresourceinvestigation,emergency operationsupport,sustainableoceanenvironmentprotection,etc.Thesechallenges and novel applications spur a surge of interest in the research and development of autonomous marine mechatronic systems. Autonomous marine vehicles (AMVs) arethosesystemswiththeintegrationofelectronics,mechanics,advancedsensors andactuators,andcontrolalgorithmsappliedinthemaritimeenvironment.AMVs playanessentialroleinperformingmarinetaskssuchasoffshoreinspections,ocean exploration,andtransportation,whichrequireeffectiveandreliablemotioncontrol systems.MotioncontroldesignforAMVsystemsischallenging,especiallywhenthe physicalconstraints,thecontrolperformance,thesafety,andtherobustnessofthese systemsneedtobesimultaneouslyconsidered.Inthepastseveraldecades,various advancedcontrolstrategieshavebeenproposedforsomerepresentativesystemsin thisarea.Amongtheseapproaches,modelpredictivecontrol(MPC)isanemerging andpromisingcontrolscheme. The essential idea of MPC can be traced back to the 1960s. Until the 1980s, MPCwassuccessfullyappliedtoindustrialapplicationsindealingwithconstrained multi-variable control problems, especially in process control. Since then, MPC has started to emerge as an advanced control technique and to attract increasing interest concurrently both from academia and industry. MPC has received consid- erable research attention and achieved successful applications in both practical and theoretical dimensions. Its success is essentially and mainly due to its ability forhandlingphysicalconstraints,multi-variableinteractions,optimalperformance, externaldisturbances,parametricuncertainties,network-inducedconstraints,cyber attacks,etc.,underaunifiedframework.Ingeneral,whenapplyingtheMPCframe- work to mechatronic systems, there are some particular issues and challenges to overcomeintheanalysisanddesign.Firstly,weneedtocomprehensivelyconsider the practical issues, physical constraints, system characteristics, control objectives andtasks,etc.Secondly,weshouldchoosethesuitableMPCtechniqueoraseamless vii viii Preface integrationofseveralMPCtechniquesforfulfillingthespecifictask.Thirdly,weneed torigorouslyanalyzethetheoreticalpropertiesoftheresultingclosed-loopsystem byapplyingtheMPCstrategy.Finally,weshouldconsiderthepracticalimplemen- tation of the MPC strategy in real-time control systems. It is underscored that the MPCstudiesarenotmerelydrivenbytheoreticalcuriosity,butalsoaretantamount toenhancedperformancewithpracticalvalue. There is also increasing awareness from the AMVs community of the need for accommodating the physical (states and control input) constraints and achieving certain optimal performance. This sets the stage for the emergence of a new application-oriented research—MPC for the motion control of AMVs—aiming at supplyingtheMPCalgorithmdesign,analysisofthetheoreticalproperties,andthe implementation guidance. Our intention through this book is to provide a timely accountaswellasanintroductoryexposuretosomeofourmaindevelopmentsinthe applicationofMPCforAMVs.Thebookwillpresentsomedesignandanalysisof MPC-basedmotioncontrolalgorithmsforthenonlinearAMVsystemsfordifferent marinemissions.Wealsohopethatthebookmayfurthersparknewideas,thoughts, anddirectionsonthesubject. Thebookconsistsofninechapters.Chapter1providesacomprehensiveoverview ofthemotioncontrolofAMVsandabriefintroductiontoMPC.Chapter2reviews the modeling of the AMV system and explores some important properties asso- ciated with the system model. Chapter 3 studies the combined path planning and trajectory-trackingcontrolofanautonomousunderwatervehicle(AUV).Aunified receding horizon optimization framework is developed with a novel spline-based pathplanningmethodandthenonlinearmodelpredictivetrackingcontrollerdesign. Chapter4developsaLyapunov-basedMPCapproachforthedynamicpositioning and trajectory-tracking control of the AUV. Chapter 5 solves the path-following problem of the AUV by developing a multi-objective MPC framework. Chapter 6 focuses on the formation-tracking control problem of multi-AUV systems with externaldisturbances.AdistributedLyapunov-basedMPCalgorithmisdevelopedto solvethisproblem.Chapter7appliesarobustdistributedmodelpredictiveplatooning controlapproachforagroupofheterogeneousautonomoussurfacevehicles(ASVs) with input constraints and bounded external disturbances. Chapter 8 investigates how to reduce the heavy computational burden and increase the speed of solving the formulated optimization problem, aiming to pave the way toward the practical application of MPC to AMVs in real time. Two novel numerically efficient algo- rithms,namelythemodifiedC/GMRES(continuation/generalizedminimalresidual method)algorithmandthedistributedNMPCalgorithm,arepresentedtoreducethe computationalcomplexity.Chapter9summarizestheworkinthisbookandprovides some potential future research directions. Moreover, for it to be self-contained, a detaileddescriptionofthesoftwareimplementationandsimulationsisprovidedin theappendix. Preface ix Acknowledgements The authors would like to thank all those who have helped in accomplishing this book. The writing of this book as well as our research on the application of MPC for AMVs has benefited greatly from interactions and collaborations with many colleagues.WewishtoexpressourheartfeltgratitudetoDeminXu,WeishengYan, Mingyong Liu, Huiping Li, Jian Gao, Changxin Liu, Brad Buckham, and Colin Bradley,withwhomwehaveenjoyedpleasantandfruitfulcollaborations.In1993, the first author started his graduate studies under the supervision of Prof. Demin XuintheSchoolofMarineScienceandTechnologyatNorthwesternPolytechnical University (Xi’an, China). The second and the third authors also graduated from the same school and then joined the Applied Control and Information Processing Laboratory (ACIPL) at the University of Victoria to pursue their Ph.D. degrees. ThreeofthemtogetherdedicatethebooktoProf.Xuasawayofexpressingtheir deepgratitudeandindebtedness,forguidingthemintotheamazingworldofmarine engineeringandcontrolengineering.SupportfromNaturalSciencesandEngineering Research Council (NSERC) has been very helpful and is greatly acknowledged. Finally, all the authors would like to greatly express special thanks to their family members. Victoria,Canada YangShi Ottawa,Canada ChaoShen Victoria,Canada HenglaiWei Victoria,Canada KunwuZhang