ebook img

Optimal control problems arising in forest management PDF

141 Pages·2019·0.973 MB·English
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Optimal control problems arising in forest management

SPRINGER BRIEFS IN OPTIMIZATION Alexander J. Zaslavski Optimal Control Problems Arising in Forest Management 123 SpringerBriefs in Optimization SeriesEditors SergiyButenko MirjamDür PanosM.Pardalos JánosD.Pintér StephenM.Robinson TamásTerlaky MyT.Thai SpringerBriefs in Optimization showcases algorithmic and theoretical tech- niques,casestudies,andapplicationswithinthebroad-basedfieldofoptimization. Manuscripts related to the ever-growing applications of optimization in applied mathematics, engineering, medicine, economics, and other applied sciences are encouraged. Moreinformationaboutthisseriesathttp://www.springer.com/series/8918 Alexander J. Zaslavski Optimal Control Problems Arising in Forest Management 123 AlexanderJ.Zaslavski DepartmentofMathematics TheTechnion–IsraelInstituteofTechn RishonLeZion,Israel ISSN2190-8354 ISSN2191-575X (electronic) SpringerBriefsinOptimization ISBN978-3-030-23586-4 ISBN978-3-030-23587-1 (eBook) https://doi.org/10.1007/978-3-030-23587-1 ©TheAuthor(s),underexclusivelicensetoSpringerNatureSwitzerlandAG2019 Thisworkissubjecttocopyright.AllrightsarereservedbythePublisher,whetherthewholeorpartof thematerialisconcerned,specificallytherightsoftranslation,reprinting,reuseofillustrations,recitation, broadcasting,reproductiononmicrofilmsorinanyotherphysicalway,andtransmissionorinformation storageandretrieval,electronicadaptation,computersoftware,orbysimilarordissimilarmethodology nowknownorhereafterdeveloped. Theuseofgeneraldescriptivenames,registerednames,trademarks,servicemarks,etc.inthispublication doesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfromtherelevant protectivelawsandregulationsandthereforefreeforgeneraluse. Thepublisher,theauthors,andtheeditorsaresafetoassumethattheadviceandinformationinthisbook arebelievedtobetrueandaccurateatthedateofpublication.Neitherthepublishernortheauthorsor theeditorsgiveawarranty,expressorimplied,withrespecttothematerialcontainedhereinorforany errorsoromissionsthatmayhavebeenmade.Thepublisherremainsneutralwithregardtojurisdictional claimsinpublishedmapsandinstitutionalaffiliations. ThisSpringerimprintispublishedbytheregisteredcompanySpringerNatureSwitzerlandAG. Theregisteredcompanyaddressis:Gewerbestrasse11,6330Cham,Switzerland Preface The growing importance of optimal control has been recognized in recent years. This is due not only to impressive theoretical developments but also because of numerous applications to engineering, economics, life sciences, etc. This book is devoted to the study of a class of optimal control problems arising in forest management.Theforestmanagementproblemisanimportantandinterestingtopic inmathematicaleconomics thatwasstudiedbymanyresearchersincludingNobel laureate P.A. Samuelson [68]. As usual, for this problem, the existence of optimal solutions over infinite horizon and the structure of solutions on finite intervals are underconsideration.Inourbooks[84,86],westudyaclassofdiscrete-timeoptimal control problems which describe many models of economic dynamics except for themodelofforestmanagement.Thishappensbecausesomeassumptionsposedin [84, 86], which are true for many models of economic dynamics, do not hold for the model of forest management. By this reason, the forest management problem is not a particular case of general models of economic dynamics and is studied separatelyintheliterature.Inthisbook,westudytheforestmanagementproblem usingtheapproachintroducedandemployedinourresearch[80,81,83].Namely, weanalyzeaclassofoptimalcontrolproblemswhichcontains,asaparticularcase, theforestmanagementproblem.Forthisclassofproblems,weshowtheexistence of optimal solutions over infinite horizon and study the structure of approximate solutionsonfiniteintervalsandtheirturnpikeproperties,thestabilityoftheturnpike phenomenon, and the structure of approximate solutions on finite intervals in the regionsclosetotheendpoints. InChap.1,weprovidesomepreliminaryknowledgeonturnpikeproperties.The forestmanagementproblemisdiscussedinChap.2,whichalsocontainsexistence resultsforinfinitehorizonproblems.InChap.3,weestablishtheturnpikeproperties ofapproximatesolutions.Chapter4containsgenericturnpikeresults.Weconsider aclassofoptimalcontrolproblemswhichisidentifiedwithacompletemetricspace of objective functions and show the existence of a G everywhere dense subset of δ themetricspace,whichisacountableintersectionofopeneverywheredensesets, suchthattheturnpikepropertyholdsforanyofitselement.Chapter5isdevotedto v vi Preface thestudyofthestructureofapproximatesolutionsonfiniteintervalsintheregions closetotheendpoints.InChap.6,weagainconsidertheforestmanagementproblem andshowthattheresultsofChaps.3and5aretrueforit. RishonLeZion,Israel AlexanderJ.Zaslavski October30,2018 Contents 1 Introduction .................................................................. 1 1.1 ConvexDiscrete-TimeProblems...................................... 1 1.2 TheTurnpikePhenomenon............................................ 8 2 InfiniteHorizonOptimalControlProblems ............................. 11 2.1 TheForestManagementProblem..................................... 11 2.2 InfiniteHorizonProblemsWithoutDiscounting..................... 16 2.3 AuxiliaryResults ...................................................... 21 2.4 ProofsofTheorems2.6and2.7....................................... 23 2.5 ProofofTheorem2.9.................................................. 24 2.6 ProofofTheorem2.10 ................................................ 26 2.7 InfiniteHorizonProblemswithDiscounting......................... 27 2.8 AnAuxiliaryResultforTheorem2.13............................... 28 2.9 ProofofTheorem2.13 ................................................ 30 2.10 AnApplicationtotheForestManagementProblem................. 30 3 TurnpikeProperties ......................................................... 33 3.1 PreliminariesandMainResults....................................... 33 3.2 AuxiliaryResults ...................................................... 38 3.3 ProofofTheorem3.2.................................................. 44 3.4 ProofofTheorem3.3.................................................. 45 3.5 ProofofTheorem3.4.................................................. 47 3.6 ProofofTheorem3.5.................................................. 50 3.7 ProofofTheorem3.6.................................................. 51 3.8 StabilityResults ....................................................... 53 3.9 AgreeablePrograms................................................... 58 4 GenericTurnpikeProperties ............................................... 63 4.1 Preliminaries........................................................... 63 4.2 EquivalenceoftheTurnpikeProperties .............................. 64 4.3 GenericResults........................................................ 67 vii viii Contents 5 StructureofSolutionsintheRegionsClosetotheEndpoints .......... 73 5.1 Preliminaries........................................................... 73 5.2 LagrangeProblems.................................................... 78 5.3 AnAuxiliaryResultforTheorem5.13............................... 83 5.4 ProofofTheorem5.13 ................................................ 85 5.5 TheFirstClassofBolzaProblems.................................... 89 5.6 AnAuxiliaryResultforTheorem5.15............................... 90 5.7 ProofofTheorem5.15 ................................................ 93 5.8 TheSecondClassofBolzaProblems................................. 98 5.9 AuxiliaryResultsforTheorem5.17.................................. 99 5.10 ProofofTheorem5.17 ................................................ 102 6 ApplicationstotheForestManagementProblem ....................... 109 6.1 Preliminaries........................................................... 109 6.2 AuxiliaryResults ...................................................... 112 6.3 TurnpikeResults....................................................... 118 6.4 GenericResults........................................................ 122 References......................................................................... 131 Index............................................................................... 135 Chapter 1 Introduction Thestudyofoptimalcontrolproblemsandvariationalproblemsdefinedoninfinite intervals and on sufficiently large intervals has been a rapidly growing area of research [9, 10, 16, 18, 24, 37, 43, 48, 57, 59, 66, 76, 84–86] which has various applicationsinengineering[1,44,87],inmodelsofeconomicgrowth[2,5,14,20, 23,28,38–40,47,52–55,60,61,63,65,67,68,70,77,80,81,83],inthegametheory [29,32,42,74,82],ininfinitediscretemodelsofsolid-statephysicsrelatedtodis- locationsinone-dimensionalcrystals[6,71],andinthetheoryofthermodynamical equilibriumformaterials[19,45,50,51].Discrete-timeproblemswereconsidered in [7, 8, 13, 21, 26, 30, 33, 72, 73, 78, 79] while continuous-time problems were studiedin[3,4,11,12,15,17,22,25,27,31,41,46,49,56,58,62,69,75]. In this chapter we discuss turnpike properties for a class of simple convex dynamicoptimizationproblems. 1.1 ConvexDiscrete-TimeProblems Let Rn be the n-dimensional Euclidean space with the inner product (cid:2)·,·(cid:3) which inducesthenorm (cid:2) (cid:4) (cid:3)n 1/2 |x|= x2 , x =(x ,...,x )∈Rn. i 1 n i=1 Let v : Rn × Rn → R1 be bounded from below function. We consider the minimizationproblem ©TheAuthor(s),underexclusivelicensetoSpringerNatureSwitzerlandAG2019 1 A.J.Zaslavski,OptimalControlProblemsArisinginForestManagement, SpringerBriefsinOptimization,https://doi.org/10.1007/978-3-030-23587-1_1 2 1 Introduction T(cid:3)−1 v(xi,xi+1)→min, (P0) i=0 suchthat{x }T ⊂Rnandx =z, x =y, i i=0 0 T whereT isanaturalnumberandthepointsy,z∈Rn. Theinterestindiscrete-timeoptimalproblemsoftype(P )stemsfromthestudy 0 ofvariousoptimizationproblemswhichcanbereducedtoit,e.g.,continuous-time controlsystemswhicharerepresentedbyordinarydifferentialequationswhosecost integrand contains a discounting factor [43], tracking problems in engineering [1, 44, 87], the study of Frenkel-Kontorova model [6, 71], and the analysis of a long slenderbarofapolymericmaterialundertensionin[19,45,50,51].Optimization problemsofthetype(P )wereconsideredin[72,73]. 0 Inthissectionwesupposethatthefunctionv :Rn×Rn →R1isstrictlyconvex anddifferentiableandsatisfiesthegrowthcondition v(y,z)/(|y|+|z|)→∞as|y|+|z|→∞. (1.1) Weintendtostudythebehaviorofsolutionsoftheproblem(P )whenthepoints 0 y,zandtherealnumberT varyandT issufficientlylarge.Namely,weareinterested tostudyaturnpikepropertyofsolutionsof(P )whichisindependentofthelength 0 of the interval T, for all sufficiently large intervals. To have this property means, roughly speaking, that solutions of the optimal control problems are determined mainly by the objective function v and are essentially independent of T, y, and z. Turnpikepropertiesarewellknowninmathematicaleconomics.Thetermwasfirst coinedbySamuelsonin1948(see[67])whereheshowedthatanefficientexpanding economywouldspendmostofthetimeinthevicinityofabalancedequilibriumpath (alsocalledvonNeumannpath).Thispropertywasfurtherinvestigatedforoptimal trajectoriesofmodelsofeconomicdynamics(see,forexample,[47,53,65]andthe referencesmentionedthere).Manyturnpikeresultsarecollectedin[76,84,86]. Inordertomeetourgoalweconsidertheauxiliaryoptimizationproblem v(x,x)→min, x ∈Rn. (P ) 1 Itfollowsfromthestrictconvexityofvand(1.1)thattheproblem(P )hasaunique 1 solutionx¯.Let ∇v(x¯,x¯)=(l ,l ), (1.2) 1 2 wherel , l ∈ Rn.Sincex¯ isasolutionof(P )itfollowsfrom(1.2)thatforeach 1 2 1 h∈Rn, (cid:2)l ,h(cid:3)+(cid:2)l ,h(cid:3)=(cid:2)(l ,l ),(h,h)(cid:3) 1 2 1 2 = lim t−1[v(x¯ +th,x¯ +th)−v(x¯,x¯)]≥0. t→0+

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.