Crowds in equations an introduction to the microscopic modeling of crowds 1st edition bertrand maury

Page 1


Visit to download the full and correct content document: https://textbookfull.com/product/crowds-in-equations-an-introduction-to-the-microscop ic-modeling-of-crowds-1st-edition-bertrand-maury/

More products digital (pdf, epub, mobi) instant download maybe you interests ...

The Madness of Crowds Gender Race and Identity Douglas Murray

https://textbookfull.com/product/the-madness-of-crowds-genderrace-and-identity-douglas-murray/

The Madness of Crowds Gender Race and Identity 1st Edition Douglas Murray

https://textbookfull.com/product/the-madness-of-crowds-genderrace-and-identity-1st-edition-douglas-murray/

Moon Marrakesh Beyond Day Trips Local Spots Strategies to Avoid Crowds Lucas Peters

https://textbookfull.com/product/moon-marrakesh-beyond-day-tripslocal-spots-strategies-to-avoid-crowds-lucas-peters/ Crowds and Sultans Urban Protest in Late Medieval Egypt and Syria Amina Elbendary

https://textbookfull.com/product/crowds-and-sultans-urbanprotest-in-late-medieval-egypt-and-syria-amina-elbendary/

Stochastic Differential Equations: An Introduction with Applications in Population Dynamics Modeling 1st Edition

https://textbookfull.com/product/stochastic-differentialequations-an-introduction-with-applications-in-populationdynamics-modeling-1st-edition-michael-j-panik/

Ordinary differential equations : an introduction to the fundamentals Second Edition. Edition Howell

https://textbookfull.com/product/ordinary-differential-equationsan-introduction-to-the-fundamentals-second-edition-editionhowell/

An Invitation to Applied Mathematics Differential Equations Modeling and Computation 1st Edition Carmen Chicone

https://textbookfull.com/product/an-invitation-to-appliedmathematics-differential-equations-modeling-and-computation-1stedition-carmen-chicone/

An Introduction to Modeling Neuronal Dynamics 1st Edition Christoph Börgers (Auth.)

https://textbookfull.com/product/an-introduction-to-modelingneuronal-dynamics-1st-edition-christoph-borgers-auth/

Introduction to Complex Theory of Differential Equations Savin

https://textbookfull.com/product/introduction-to-complex-theoryof-differential-equations-savin/

Crowds in Equations

An Introduction to the Microscopic Modeling of Crowds

Advanced Textbooks in Mathematics

Print ISSN: 2059-769X

Online ISSN: 2059-7703

The Advanced Textbooks in Mathematics explores important topics for postgraduate students in pure and applied mathematics. Subjects covered within this textbook series cover key fields which appear on MSc, MRes, PhD and other multidisciplinary postgraduate courses which involve mathematics.

Written by senior academics and lecturers recognised for their teaching skills, these textbooks offer a precise, introductory approach to advanced mathematical theories and concepts, including probability theory, statistics and computational methods.

Published

Crowds in Equations: An Introduction to the Microscopic Modeling of Crowds by

The Wigner Transform by Maurice de Gosson

Periods and Special Functions in Transcendence by Paula B Tretkoff

Mathematics of Planet Earth: A Primer by Jochen Bröcker, Ben Calderhead, Davoud Cheraghi, Colin Cotter, Darryl Holm, Tobias Kuna, Beatrice Pelloni, Ted Shepherd and Hilary Weller edited by Dan Crisan

Forthcoming

Conformal Maps and Geometry by Dmitry Beliaev

Crowds in Equations

An Introduction to the Microscopic Modeling of Crowds

Bertrand Maury

Université Paris-Sud, France

Sylvain Faure

Université Paris-Sud & CNRS, France

Published by

World Scientific Publishing Europe Ltd.

57 Shelton Street, Covent Garden, London WC2H 9HE

Head office: 5 Toh Tuck Link, Singapore 596224

USA office: 27 Warren Street, Suite 401-402, Hackensack, NJ 07601

Library of Congress Cataloging-in-Publication Data

Names: Maury, Bertrand, author. | Faure, Sylvain, 1976– author.

Title: Crowds in equations : an introduction to the microscopic modeling of crowds / by Bertrand Maury (Université Paris-Sud, France), Sylvain Faure (Université Paris-Sud & CNRS, France).

Description: New Jersey : World Scientific, 2018. | Series: Advanced textbooks in mathematics | Includes bibliographical references and index.

Identifiers: LCCN 2018013165 | ISBN 9781786345516 (hc : alk. paper)

Subjects: LCSH: Mathematical analysis. | Mathematical models. | Communication in mathematics.

Classification: LCC QA401 .M448 2018 | DDC 511/.8--dc23

LC record available at https://lccn.loc.gov/2018013165

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.

Copyright © 2019 by World Scientific Publishing Europe Ltd.

All rights reserved. This book, or parts thereof, may not be reproduced in any form or by any means, electronic or mechanical, including photocopying, recording or any information storage and retrieval system now known or to be invented, without written permission from the Publisher.

For photocopying of material in this volume, please pay a copying fee through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA. In this case permission to photocopy is not required from the publisher.

For any available supplementary material, please visit http://www.worldscientific.com/worldscibooks/10.1142/Q0163#t=suppl

Desk Editors: V Vishnu Mohan/Jennifer Brough/Koe Shi Ying

Typeset by Stallion Press

Email: enquiries@stallionpress.com

Printed in Singapore

Foreword

Buildingbridges,remarkingsimilarities,crossingmethodsareessential drivingforcesofthescientificactivity.Sinceancienttimes,physicsand mathematicshavebeeninterwoven.Today,althoughtheamountofknowledgemakesitessentiallyimpossibletohaveaglobalviewontheseso-called hardsciences,interactionsremainfruitful.Ontheotherhand,socialsciencesare—atfirstsight—unconcernedwiththisthinkingsystem:they areinterestedinamultitudeofbehaviourswhichseemtoescapethedeterministiclawsofphysicsandanymathematicalformulation.

Asmallrevolutionofthelastdecadehasbeentoremovethisbarrier toallowanewresearchfieldtoemerge.Ifoneobservestypicalbehaviours (atthescaleofmanyindividuals)andthatonecanidentifyparameters influencingthesebehaviours,thenthereisnoreasonthatthemathematical languagecannotdescribethem!Ofcourse,theconnectionisnotsoeasyand thisnewfieldisstillinitsinfancy,notalwaysconsideredseriouslybyits elders.

Inthisbook,intendedtograduatestudentsandresearchersin mathematics,SylvainFaureandBertrandMauryinviteustodiscoverthe challengesandthefirstsuccessesofmathematicsappliedtosocialsciences. Asapreamble,theyclearlyexplainthedifficultiesoftheexercise,due,in particular,tothefreedomofindividualsandtothedecisionprocesseswhich areneithersymmetric,norinterchangeable.

Thebookcontinueswiththerigorousanalysisofsomemodels,essentiallyatthemicroscopicscale,whichserveasmathematicalprototypesand exhibitinterestingphenomenologies.

Startingfromthispicture,theauthorsproposeeventuallytoextract theminimalelementswhichshouldb econtainedinamathematicalmodel inordertoreproducesometypicalandsometimesparadoxicalproperties ofcrowdmotions:“Faster-is-Slower”effect,“Stop-and-Go”waves,andfluidizingeffectsofanobstacle.Thisisfascinating!Aquickandeasyread, whichmakesmewanttolearnmore.

EcoleNormaleSup´erieuredeLyon& Acad´emiedesSciences,France

3.SocialForceModel,NativeandOverdampedForms37

5.CellularAutomata83

5.1CellularAutomata:GeneralPrinciples...........84

5.2Algorithms..........................85

5.3Variations,Extensions....................92

5.4CellularAutomata,MathematicalIssues..........93

6.CompartmentModels97

6.1CompartmentModels:ToyVersions andGeneralSetting.....................97

6.2NumericalSolution......................101

6.3Extensions...........................102

6.4NumericalIllustration....................104

6.5MathematicalFramework:ACascade ofGradientFlows.......................104

7.TowardMacroscopicModels111

7.1One-DimensionalMacroscopicTrafficModel........112

7.2Two-DimensionalModels..................115

7.3GranularModels:HardCongestion.............117

7.4Micro–MacroIssues......................123

7.5AlternativeMacroscopicModels...............125

8.ComputingDistancesandDesiredVelocities127

8.1ShortestPathProblemonaGraph.............130

8.2ShortestPathonaDomain:TheEikonalEquation....132

8.3Non-homogenousDomains,VariousExtensions......135

8.4ShortestPathsinaDynamicEnvironment.........139

8.5AlternativeWaytoComputeDesiredVelocities......142

8.6Illustrations..........................143

9.Data,ObservablePhenomena145

9.1Diameters...........................145

9.2Proxemics,InterpersonalDistances,Density........146

9.3ConeofVision........................148

9.4PedestrianSpeed,FundamentalDiagram..........148

9.5DoorCapacity........................151

9.6CapacityDropPhenomenon.................151

9.7Faster-is-SlowerEffect....................152

9.8InfluenceofanObstacle...................154

9.9Stop-and-GoWaves......................157

9.10FurtherConsiderationsonHumanBehavior........158

10.AWiderLookonCharacteristicPhenomenainCrowds161

10.1Faster-is-SlowerEffect....................161

10.2FluidizingEffectofanObstacle...............168

10.3Damping,Propagation,andStop-and-GoWaves......170

This page intentionally left blank

Chapter1 Introduction

1.1.FromPassivetoActiveEntities

Themodelingofparticlesystemshasraisedaconsiderableactivityinthe lastcenturies.Physicists,mathematicians,andmorerecentlycomputerscientists,havejoinedtheireffortstoformalizethelawswhichgovernthe motionofparticlesininteractionwithoneanother.Theybuilttheoreticalframeworksandnumericaltoolstoevolvethe“bigpicture”,i.e.induce generalrulesormacroscopicequationswhichwouldmakeitpossibleto describethebehavioroftheconsideredsystematalargescale,beyondthe individualdestiniesofitscomponents.

Sinceafewdecades,scientistsfromvariousdomainshaveextendedthe approachtosystemsof“activeentities”,birdflocks,fishschools,crowdsof mammalsorinsects,and,evenfurtherapartfrompassiveparticles,walkingordrivinghumanbeings.Partoftheframeworksdevelopedforphysicalsystemscanbestraightforwardlytransposedtothisnewsituation.In particular,thekinematicmodelingwillconsistinrepresentingamoving crowdbyatime-varyingvectorwhichcontainsthepositionsofitsindividuals.Nowassumingoneisabletomodelthewillofanindividualand itsinteractionwithothersinanequationinvolvingthepositionsandits derivatives,themodeltakestheformofasystemofdifferentialequations, forwhichatremendousamountoftheoreticalandnumericaltoolshave beendeveloped.Yet,themodelingoflivingentitiespresentsparticularfeatureswhichmakeitverypeculiaramongotheractivitiesinparticlesystem modeling.

(1)(Outofequilibriumthermodynamics)Thefactthattheconsideredentitiesareactivefromthe dynamicalpointofview(theyare capableofusingtheirinternalenergyfortheirownmotion)rulesout thepossibilitytoexpectsomekindofthermalizationoftheoverall system.

(2)(Decisionprocesses)Humanbeingsarealso active inadecisional sense.Whereaspassiveparticlesobeymechanicallaws,andpreprogrammedautomataevolveaccordingtopredeterminedrules,human beingshavetheabilitytodesigntheirowninstantaneousstrategies. Suchchoicesmaybebasedonafullorpartialknowledgeoftheirenvironment(e.g.positionsandpossiblyvelocitiesofneighbors),andalso triggeredbyemotionalfactors(impatience,stress,justifiedorartificial panic, ... ).

(3)(Asymmetricinteractions)Asmentionedinthepreviouspoint, humanagentstakedecisionsaccordingtotheirinstantaneousknowledge.Inthecontextofcrowdmotion,thelattermostlycomefrom vision.Sincevisionisoriented(pedestrianstypicallyvisualizeacertain coneaboutthedirectiontheyheadto),theinfluencerelationswithin acrowdarealsooriented,whichkillsthesymmetryofinteractions. Althoughthenotionof socialforce iscommonlyusedtoaccountfor interactions,itmustnotbeforgottenthatthoseso-calledforcesmay ruleouttheLawofAction–Reactionwhichgovernsthedynamicsof passiveparticles.

(4)(Mesoscopicscale)Someimportantfeaturesincrowdmotiondevelop atamesoscopicscale,i.e.theyappearandmustbedescribedatascale thatisnotinfinitelylargewithrespecttothemicroscopicscale(i.e.the sizeofanindividual).Asweshallsee,itconcernsforexample“Stopand-Go”wavesinaqueue,thewavelengthofwhichmerelyextendsto afewindividuals(tobecomparedtoaudiblesoundwavesinagas,the typicallengthofwhichscaleslike108 timestheinter-particledistance). Inadifferentcontext,theevacuationofaroomstronglydependson whathappensintheveryneighborhoodoftheexitdoor,thesizeof whichscaleslikepeoplediameter.

(5)(Non-interchangeableentities)Althoughmostmodelspresentedin thisbookwillbebasedonthesimplifyingassumptionthatindividuals behaveinasimilarway,itshouldnotbeforgottenthatrealcrowdsare madeofindividualswithdifferentpersonalities(intermsofaggressivity,politeness,abilitytodevelopstrategies,tendencytocooperate, ).

Assuch,thestateofthesystemcannotbedefineduptopermutations, likeinthecaseofgasparticles.

1.2.BasicsonCrowdMotionModeling

Weshallessentiallyrestrictourselvesinthisbooktomicroscopicmodels. Thosearebasedonanindividualtrackingofindividuals,andtheyareby nature Lagrangian:eachvariablewillbeattachedtoagivenindividual. Mostmicroscopicmodelsrelyontwomainingredients:

(1)Definingindividualtendencies,i.e.choosingtractablerulestodetermine,ateachinstant,whatanindividualwoulddoiftheywerealone. Notethatsuchanissuemaynotberelevantinsomesituations,e.g. forachildfollowinghisparents,butweshallgenerallyconsiderthat theatomicentityofthemodelisaresponsiblepersonwithanown intention.Inthecaseofabuildingevacuation,whichwillbeusedasan archetypalexampleinthisbook,suchanintentionisclearlydefined: reachtheclosestexitinashorttime.

(2)Interactionrules:howisapedestrianinfluencedbyotherpedestrians intheirneighborhood?Thecoreofeachmodelreliesinaccounting forthoseinteractions,whichmaybeofvarioustypes.Inthisbookwe shallfocusontwotypesofinteraction:physicalinteractionbetweentwo peopleincontact,orsocialtendencytopreserveacertaindistanceto neighbors.Thefirsttypeofinteractionispurelymechanical,andfits intheclassicalframeworkofgranularmechanics(theterm grains is preferredtothatof particle whenfinite-sizeeffectsaresignificant).The secondtypeischaracteristicoflivingentities,sinceitisnottriggeredby aphysicalcontact,butratherimplementsacomplexcognitiveprocess whichisinitiatedbytheperception(typicallyvision)ofneighbors.As such,itmaydisobeytheLawofAction–Reaction.

Thewaythoseingredientsareencodedinthevariousmodelsdependsonthe typeof representation whichischosen,asdetailedinthenexttwosections.

Individualtendencies

Asforindividualtendencies,thecorenotionwillbethatof desiredvelocity, alloverthisbook.Letusmakeitclearthatthisnotionissomewhatambiguous,sinceonemayconsiderthattheinstantaneousdesireofacivilized

pedestrianaccountsforpeopleintheirneighborhood.Ourchoice,following thatofmostauthors,willbetofavortheselfishmeaningof“desired”:the desiredvelocity,whichwemayalsocallspontaneousvelocity,isthevelocity thatapersonwouldhave iftheywerealone.Itthereforecorrespondstoan instantaneoustendencyofachievingagivengoal(e.g.exitingabuildingin fire)inapurelyselfishway.Intheone-dimensionalsetting(seeChapter2), pedestriansareassumedtowalkonebehindanother,allheadingtothe samedirection.Underthoseassumptions,thedesiredvelocityturnsoutto beadesired speed,or freespeed.Inthetwo-dimensionalsetting,thedesired directionmustalsobeprescribed.Chapter8isdedicatedtothequestion ofdeterminingthedesiredvelocityfieldinvariouscontexts.Weshallfavor scenariosoftheevacuationtype,whichmakesitpossibletoproperlysetthe problemwithouttoomuchpsychologicalconsiderations:thedesiredvelocityofagentswillbebuiltasbestsuitedtoachievethecommongoal,that istoexitthebuildingasfastaspossible.Ingeneral,thisvelocitywillbe defined(uptoamultiplicativeconstant)astheoppositeofthegradientof aquantitythatapedestriantendstominimize.Ingeneral,thisquantity hasavocationtorepresentapedestrian’sdissatisfaction,likethedistance totheexitforanevacuation.

Theactualencodingoftheindividualtendencydependsonthemodel representation.Intheinertialsocial forcemodel(Chapter3),itisencoded asarelaxationterm:thepedestrianb ehaveslikeaninertialparticleina viscousfluid,thevelocityofwhichisthedesiredvelocity.Viscousfrictional forcestendtohavethepedestrian/particlemoveatthesamevelocitythan theunderlyingfluid.Thenon-inertialversionofthismodel(Section3.2in thesamechapter)correspondstothecaseofaninfinitelyviscousfluid:the particlebehaveslikeapassivetrackerwhichfollowsthemotionoftheunderlyingfluid.Thedifferentialsystemisthensetonthepositionsonly,and thedesiredvelocityisthemainforcingtermwhichconditionstheactual velocity(correctionsduetointeractionswillbeaddressedbelow).Inthe granularapproach(Chapter4),thesetofdesiredvelocityfieldisconsideredassomesortof attempted velocityfield,whichisprojectedtotheset ofadmissiblevelocities(i.e.velocitiesthatrespectthenon-overlappingconstraint).InthecontextofCellularAutomata(Chapter5),pedestriansare consideredastokensoccupyingthecellsofacartesiangrid,andtheevolutionprocessisofastochasticnature.Inthiscontext,individualtendencies areimplementedasbiastothehoppingprobabilitiestowardneighboring cell.Insomeway,CAimplementinastochasticwaytheassumptionthat thedesiredvelocityheadsalongtheoppositeofthegradientofaquantity

thatistobeminimized.InthecontextofCA,thisquantityiscalledthe floorfield,butitplaystheexactsameroleasthedissatisfactionalready mentioned.

Interactionrules

Itwouldbequitepresumptuoustosuggestthatallimaginablesocialinteractionscanbeencodedinequations.Weshallthereforerestrictourselvesin thisbooktosimpletypesofinteractions,inparticularthesocialtendency tomaintainacertaindistancefromneighbors,togetherwiththephysical non-overlappingrule(twoindividualsmaynotoccupythesamespaceat thesametime).

Intheone-dimensionalsetting(Chapter2),withpedestrianswalking onebehindanother,itcanbeconsideredthatwalkerstendtopreserve acertaindistancetothewalkerbehindinordertoavoidacollisionin thecasewherethelattersuddenlystops.Expressedinareciprocalway, thistendencywillbeencodedinassumingthatthespeedisanincreasing functionofthedistancetothenextwalker.

Inthegeneral,two-dimensionalsetting,eveninthestaticorquasistaticsituation,thetendencytomaintainacertaindistancefromneighbors willbeencodedinvariousways,dependingontherepresentation.Inthe so-calledsocialforcemodel(anditsextensions),presentedinChapter3, thistendencyismodeledbyarepulsiveforce,whichpossiblyrulesoutthe LawofAction–Reaction(ifconeofvisionsareaccountedfor).

Anotherapproachconsistsinrepresentingindividualsbyrigiddisks(see Chapter4),andtorestrictinteractionstoactualphysicalcontacts.Inthis framework,interactionforcesappearasmathematicalauxiliaryvariables associatedtonon-overlappingconstraints,namelyLagrangemultipliers.

Inthecellularautomataapproach,therepresentationisdifferent(as alreadymentioned,anindividualisidentifiedtoatokenplacedinacell ofacartesiangridwhichcoversthezoneofinterest),butinteractionsare alsotreatedinahardway,bysimplyrequiringthattwoagentsarenot authorizedtoshareacommoncell.

1.3.TheMathematicalStandpoint

Sincemostpublicationsonmicroscopiccrowdmodelscomefromthecommunityofphysicistsorcomputerscientists,whilethisbooktendstoadopt amathematicalstandpoint,itismandatorytomakesomeremarksonthe

“cultural”differencesbetweenthoseapproaches.Foramathematician,a model(inthepresentcontextofcrowdmotion)consistsinanequation orasetofequationsoftwomaintypes:OrdinaryDifferentialEquations (ODEs)involveafinite(butpossiblylarge)numberofvariablesdepending ontime(seenasaone-dimensionalcontinuum),whereasPartialDifferentialEquations(PDEs)involveunknownfieldsdefinedoveracertainzone intheEuclideanspace.Suchasetofequationswillbeconsideredasa soundmodel(fromthemathematicalstandpoint)ifitpresentsapredictivecharacter,i.e.ifonecanestablishthat,giveninitialconditions(and possiblyboundaryconditionsforPDE’s),existenceanduniquenessofa solutioncanberigorouslyproven.Theterm predictive,atthisstage,does notmeanthatthesolutionproperlydescribesinapredictivewayanyreallifephenomenon,itsimplymeansthat,undersomeassumptionsandinitial conditions,themodelcan“produce”somethingthatisproperlydefined. Establishingthiswell-posednessusuallydoesnotgiveinformationonthe solutionitself(althoughsomeexistenceproofsare constructive,andcanbe usedtoeffectivelyapproximatethesolution).Exceptforveryparticular situations,thesolutionstothoseequationsdonotadmitanalyticalexpressions,sothatnumericalcomputationsarenecessary.Thosecomputations arebasedonaninitial discretization process,necessarytoreplaceinfinitedimensionalproblemsbyproblemsinvolvingafinitenumberofunknowns, whichmakesanactualcomputationtractable.ForODEs,itsimplyconsists inreplacingthecontinuoustimelinebyadiscretesetofso-called timesteps. ForPDEs,thespaceitselfisdiscretized,i.e.thedomainisdecomposedonto afinitenumberofsmallcells,towhichunknownsareassociated.Parameterspertainingtothisdiscretizationprocesshavenomeaningintermsof modeling.Inthisframework,whenacomputationispresented,itismeant (sometimesimplicitly)thatthecomputedsolutionsthatarepresentedare closetotheexactsolution.By“close”wemeanthatitpresentsthesame features,uptobeingindistinguishableintheeyedistance.Inparticular,it requiresthatreducingforinstancethetimestepshouldnotsignificantly affectthecomputedsolution.NumericalAnalysisisthebranchofapplied mathematicsdedicatedtoproperlyquantifythiscloseness.Convergence resultstypicallyensurethat,whenthediscretizationparametersgoto0,the computedsolutionconvergestotheexactoneforsomeappropriatenorm. NumericalAnalysisisnotcentralinthepresentbook,butallnumerical methodswhichweuseareofcoursecoveredbysuchconvergenceresults.

Thestandpointofphysicistsmaydiffer:themodelingphase(establishingtheequations)andthediscretizationphasearenotalwaysstrictly

separated.Inparticular,itiscommonpracticetogiveaphysicalinterpretationtowhatamathematicianwouldcalladiscretizationparameter.This differenceisnotaminorone:itreflectsadeepdifferenceinwhatisconsideredasaModel.Weshallpresentinthisbook(seee.g.Section3.1)some ingredientsaddedtoODEmodels,insuchawaythattheinducedeffect stronglydependsonthetimestep,inparticularitvanisheswhenthetime stepgoestozero.Itleadstoconsiderthistimestepasamodelingparameter,ratherthanapurediscretizationparameter.Inthiscontext,themodel isnolongerasystemofODE’s,butanessentiallydiscreteprocedure,witha timestepthatisnotintendedtogoto0.Thisphilosophycanleadtomodels whichareessentiallydiscrete,likecellularautomata(CA,seeChapter5). Thosemodelsconsistofasuccessionofdiscreteevents,namelyhoppingof particlesfromonecelltoanother,inacartesiangrid.Thisgridissimilarto discretizationmesheswhichareusedtonumericallysolvePDE’s,yetthey aredifferentinessence,becausethespacestepisnotintendedtobetaken smallerandsmaller.InthecontextofCA,itrepresentsthezonetypically occupiedbyapedestrian.Inthiscontext,althoughthealgorithmitselfmay looksimilartosomenumericalmethodsusedtosolvePDE,itisnotrelated toanycontinuousequation.

Bothapproachesareofcoursecomplementary,anditwouldnotmake anysensetorankthemintermsofrelevance,butitisoftheutmostimportancetokeepinmindthedistinctiontoappreciatewiththerightcriteria thevariousapproachesproposedintheliterature.

Domathematicsreallymatter?

Thepreviousconsiderationsraiseanaturalquestion:whatisthecontributionofmathematicsinthemodelingprocess,dotheyreallyhelpbetterunderstandthereal-lifephenomena,theunderlyingmechanismswhich makethingshappenastheydo?Thisquestionisnotpurelyrhetorical. First,hugecontributionsinsciencehavebeenmadewithouttheuseof deeporsophisticatedmathematics(excludingthebasiclanguagethatis sharedbytheso-called hardsciences ).Second,itcannotbedeniedthat somemathematicaldevelopmentsallegedlyissuedfrommodelingquestions, asinterestingastheymaybefromthestrictmathematicalstandpoint, donotbringmuchlightontheunderlyingphenomena.Letusaddhere (thispointisacommonsourceofmisunderstandingbetweencommunities)thattherigorofthetheoreticalanalysisandthenumericalanalysis, whichisusuallyafull-timejobforappliedmathematicians,doesnotprove

anythingonthequalityofthemodelintermsofrepresentingtheunderlying reality.

Weneverthelesswouldlikeheretoadvocatefortheroleofsomemathematicaldevelopmentstobetterunderstandanddescribephysicalphenomena,uptocomplexlivingsystemslikecrowds.

Existenceanduniqueness

Letusstartwiththepetthemeinmathematics,thatisprovingexistence anduniquenessofasolutiontoagivenproblem.Letusfirstsaythatuniquenessofasolutionindicatesthatnoequationismissing,giventheconsidered unknown.Moreprecisely,ittellsthemodelerthatnoequationorconstraint shouldbeadded,andeventhatnone can beadded(additionofextraequationsorconstraintswouldoverdeterminetheproblem,andingeneralrule outexistence).Existence ofasolutionis,paradoxically,moredelicateto interpret.Togetherwithuniqueness,itbringssatisfactiontothescientist inrigorouslyallowingthemtotalkabouta solution assomethingthatis properlydefined(inawell-definedsense).Buttherichercontributionto mathematicalwell-posednessanalysistotheunderstandingamodellies oftenindarkerzones,i.e.whenthingsdonotworkastheyareexpectedto. Thefactthatwell-posednesscannotbeestablished,oronlypartially(e.g. thesolutionofanevolutionproblemexistsforashorttimeperiod,but globalexistencecannotbeproven),commonlyreflectsadefectinmodeling assumptions,oremergenceofacriticalsituationthatisnotcoveredbythe model.Letusgiveanexample:inthecontextofmicroscopiccrowdmodels, individualsarecommonlyrepresented bydisks,andinteractionsbetween twoindividualsareencodedbyvectors“acting”onbothalongthelinejoiningtheircenters.Obviously,thedirectionofthislinesmoothlydependson thecenterlocationwhenbotharewellseparated,butthesmoothnessislost whenthosecentersgetcloseruptocoincide.SuchaproximityuptocontactwillruleoutthepossibilitytouseCauchy–Lipschitztheorytoensure existenceanduniquenessofaglobalsolution.Ifamodelissuchthatthis coincidenceislikelytohappen,italsomeansthatitauthorizespedestrian togoacrosseachother,sothatthetheoreticaldefectactuallyreflectsaflaw inthemodelitself.Whoisnotawareofthisproblem(whichistypically duetoabadcalibrationofrepulsionforces,whicharesupposedtoprevent fulloverlappingfromhappening)takestherisktoobtainunrealisticresults byastraightcomputationofthisill-posedmodel.Anotherdefectinwellposednesstheorymaybeduetonon-regulardesiredvelocityfields.Weshall

describe,inthecontextofbuildingevacuation,howsuchfieldscanbebuilt inanaturalway,bysimplystatingthatthelocalvelocitycorrespondsto theshortestpathtotheclosestexit(expressedinafancyway,thevelocity shallbedefinedasthe oppositeofthegradientofthegeodesicdistanceto thesetofexits).Avelocityfieldbuiltinsuchawaymaybenotregularas soonastheroomcontainsobstacles,orifthereismorethanoneexit.This willagainruleoutastraightuseofstandardtheoremstoproveexistence anduniquenessofasolution.Again,this theoreticaldefectreflectsafeature ofthemodel.Thezoneswherethefieldisnotregularcorrespondtopoints fromwhichthereistwoormorewaystoexitthebuilding.Itobviously callsforanenrichmentofthemodelintheneighborhoodofthosesingular zone,e.g.bystatingthatanindividuallocatedinthisneighborhoodshall randomlypickoneofthepossiblewaystoexitthebuilding,orwillpickthe onetheytooktoenterthebuilding.

Stabilityanalysis

Mathematicaldevelopmentsmayprovidemoredirectcontributionsto understandingemergenceofsomephenomenaintheneighborhoodofsome particularstates,and/ortoassessthedependenceofsomeobservablequantitieswithrespecttomodelparameters.Asanillustration,crowdmotion modelsinaone-dimensionalsetting(peoplewalkingbehindeachother) aredescribedinChapter2.Theyexhibitsimplestationarystates,inthe neighborhoodofwhichafullstabilityanalysiscanbecarriedout.Thisanalysisconsistsininvestigatingthespectrum(setofeigenvalues)ofthematrix associatedtothelinearizedsystemattheequilibriumpoint.Inthecomplex plane,thezoneoccupiedbythisspectrumwithrespecttotheimaginary axisdeterminesthestabilityoftheequilibriumpoint.Formodelswhichare firstorderintime,thefactthatthespectrumliesontheleft-handsideof thisaxisensuresstability.Thereciprocaloftherealpartsoftheeigenvalues correspondstocharacteristictimes,inparticularthelargestofthosetimes givestheorderofmagnitudeofthetimethatistakenbyaperturbation oftheequilibriumstatetodampbacktoequilibrium.Thecorresponding imaginarypartscanbeinterpretedastimefrequenciesofdampedoscillationsundergonebyindividualsafterthesystemhasbeenperturbed.

Morestrikingly,thesameanalysisappliedtotheinertial(ordelayed) versionofthismodelwillexhibitnativeinstabilities,i.e.spontaneousemergenceofperturbationswhichtendtopropagateupstream(Stop-and-Go waves).

Identifyingmechanisms,abstractcauses

Thepresentbookcontainsvariousmodelsofcrowdmotion,mostlyissued frommicroscopicconsiderations:thosemodelsarebuiltbyexpressingthe behaviorofapedestrian,madeofindividualtendenciestogetherwithingredientstomimicinteractionswithotherpedestrians.Ontheotherhand, somecollectivephenomenahavebeen experimentallyobserved,someof whichareparadoxical,liketheCapacityDrop(CD)phenomenon,the Faster-is-Slower(FiS)effect,andthe fluidizingroleofanobstacle.Determiningwhetherornotagivenmodeliscapable1 ofreproducingthosephenomenahavegivenrisetoanimportantliterature(seeChapter9).Yet,the issueregardingwhichparticularingredientsinthemodelmakeitableto reproduceaphenomenonisrarelyaddressed.

Mathematicscanhelpbetterunderstandtheveryroleplayedbyvariousingredients,therebysheddingsomelightonthepossiblecausesofthe observedphenomena.Inthisspirit,afullchapter(Chapter10)isdedicated tothisapproach,whichwemaydesignateas“abstractinversemodeling”, sinceitconsistsindeterminingwhichingredientsarerequiredforageneral modeltoreproduceagivenphenomenon.ThisisdoneinChapter10by designingsomesortsofminimalmodels,i.e.equationsbasedonfewvariables,orminimalsetsofequations,whichmakeitpossibletorecoverthe phenomenon.Theapproachisinductive,anddoesnotcommonlyleadtoa singleanswer.Asanillustration,letusmentiontheFaster-is-Slowereffect, whichessentiallystatesthat,whenpeopleevacuatethroughanarrowexit, itmayhappenthatincreasingthevelocity(oreagerness)ofsomeofthem mayharmtheoverallevacuationprocess.Theapproachwhichwepropose makesitpossibletosingleouttwomain“ingredients”likelytoexplainthe paradoxicaleffect,eachofwhichalonebeingabletoexplainit,thatis: thenon-convexityofthesetoffeasibleconfigurations,andfrictionbetween individuals.

Modeling =Betraying?

Mostmodelsandmathematicaldevelopmentspresentedinthisbookare basedonahighlyidealizedrepresentationofthereality.Thisfeatureis

1 Letusmakeitclearthat,sincetheaforementionedphenomenonarenotsystematically observed,thenon-reproductionofoneofthoseeffectsdoesnotdisqualifyamodel.We referthereadertoChapter9foradetaileddiscussiononthismatter.

inherenttotheverycoreofmathematicalmodeling,whichfavorssimplicityandsparsityintermsofparameters.Theunderlyingprincipleisthe following:alwayselaboratethesimplestmodelwhichreproducesagiven feature.Thisprinciplehelpsinavoidingthecurseofoverfitting.2 Besides, thesobernessofamodelallowsfordirectuseofabstractmathematical results.Yet,thisoversimplificationhastobequestioned:

(1)Dooversimplifyingassumptionsdisqualifythemodelinitsabilityto predicttherealityinsomeway,toreproducecharacteristicfeatures, andtorecoverobservablequantitieswithareasonableaccuracy?

(2)Someobservablephenomenaareexplainedbyabstractmathematical developments.Thosemathematicaldevelopmentsmaynecessitatethe modeltobe“cleanandspare”.Sincetherealityisnot,doesthis approachreallymakesenseintermsofmodeling?

Thoseissuesgofarbeyondthescopeofthisbook,buttheyshould bekeptinmindtomaintainacertaincriticalthinkingwhiledeveloping newmodels.Weshallrestrictourselvesheretoshortlyinstantiatethose generalquestionsinconnectionwithcrowdmotions.Toillustratethefirst point,considertherepresentationofp edestrianastwo-dimensionalsimple geometricalobjects.Mostmodelsindeedrelyontheidentificationofpedestrianstodisks,whichobviouslydepartsfromreality.Someofthosemodels comparesufficientlywellwithrealitytoconvincethatthissimplificationis nottooharmful.Yet,ifsuchamodelismeanttobeusedinapredictive manner,inasituationinwhichthemodelhasnotbeenvalidated,there isnoguaranteeonitsabilitytoproperlyfulfillitstask.Asforthesecondpoint,letuspointtothespectralanalysiscarriedoutinChapter2 inordertoexplaintheupstreampropagationofaperturbationalonga lineofpedestrians.Thisanalysisreliesonstrongassumptionsintermsof evennessofthepedestrians.Thoseassumptionsareruledoutassoonas oneconsidersthatpedestriansmayslightlydifferintermsofbehavior,so thattheoveralltheory(basedonthefactthatacertainmatrixis notdiagonalizable)becomesinapplicable.Weinvestigate(seep.27andfollowing) thepossibilitytoextendtheapproachdedicatedtothehomogeneousmodel tothemorerealisticcaseofnon-homogeneouspopulations.Thisillustrates thenecessitytocomplementthetheoreticalstudyofover-simplifiedmodels

2 Amodelwithtoomanyparameters,asdisrespectfulasitmaybeoftheunderlying reality,islikelytoreproduceanytargetproperty.

byaninvestigationofthe robustness oftheconsideredmodels.Thisinvestigationaimsatcomfortingtheoverallmodelingapproachbycheckingthat thecleanandsparemodelreallycarriestheessentialfeaturesofthegeneral situation.

1.4.HowtoUsethisBook?

Thevariouschaptersofthisbookhaveb eenconceivedasself-contained,and thebookisnotbuiltinaprogressiveway.Tobemoreprecise,Chapters2–6, presentvariousindependentapproaches,allbasedonamicroscopicdescriptionofcrowds.Eachofthosechapterscanbereadatfirst.

Chapter7differsfromthepreviousones,sinceitproposesanintroductiontomacroscopicmodels,itmakessensetoreaditafterhavinga fairknowledgeinmicroscopicaspects.Itmayalsobeskippedbyreaders decidedtosticktomicroscopicapproaches.

Chapter8ismoretechnicalandtransverse:itisfocusedoncomputationalaspects.Itisinparticulardedicatedtoactualcomputationofdesired velocityfieldsinnon-trivialsituations.Sincethistechniqueisneededbyall approaches,understandingitismandatoryforreaderswhomightwantto applyanyoftheproposedmodelstoreal-lifesituations(forinstancein complexbuildings).

Chapter9isthelessmathematicallyoriented:itisentirelydedicatedto experimentalevidence,measurabledata,andobservedphenomena.

Chapter10isthemostexoticone,itcontainsfreemathematicaldevelopmentstoinvestigatethepossiblecauses(intheabstractsenseof model features)ofsomeobservedphenomenaincrowdmotions.

Westronglyencouragereaderstocomplementthetheoreticalstudyof modelsbyperformingtheirowncomputations.Numericalalgorithmsto numericallysolvemostmodelsinthisbookareproposedinthePython package cromosim,developedbytheauthors.3

3 Pythonprogramsusedinthisbookcanbeobtainedat http://www.cromosim.fr

Chapter2

One-Dimensional MicroscopicModels

Thischapterisrestrictedtoone-dimensionalmotions:pedestriansare assumedtowalkonalinetowardacommondirection.Section2.1isdedicatedtotheso-calledFollow-the-Leadermodel,whichreliesontheassumptionthattheinstantaneousvelocityofanindividualisafunctionofthe distancetothenextindividual.

Section2.2isdedicatedtoaricherversionofthismodel,including delayedreactionorinertialeffects.

2.1.Follow-the-LeaderModel

Wepresentherethebasicformofthe Follow-the-Leader(FTL)model, someextensionswhichhavebeenproposed,andweaddresssometheoretical issues.

FTLmodel:assumptionsandactualbehavior

Weconsider N +1individualswalkingonastraightline.Theirrespective positionsaredenotedby

Themodelisbasedontheassumptionthattheinstantaneousvelocityof i dependsupon xi+1 xi only.

Model2.1(FTLmodel). Let ϕ : R+ → R+ beafunctionwhichassigns aspeedtoanynon-negativedistance.Themodelsread:

Weshalldesignateby linear thismodelwhenthespeedofindividual N +1 isprescribed,and periodic thecasewhere N +1isidentifiedto1(inwhich casethelengthoftheperiodicpathhastobespecified).Thesystemisthen autonomousinthelattersituation(noexplicitdependenceupontime),and non-autonomousinthelinearcase(because xN +1 (t)hastobeprescribed inthelattersituation).

Denotingby wm > 0(m for“minimal”)theinter-individualdistance whichcorrespondstofullpacking(peopleareincontact),itisnaturalto assumethat w → ϕ(w )vanishesat wm ,increaseswith w ,andtendstoa maximalvalue U when w becomeslarge.

Anexampleofsuchafunctionis

ϕ(w )= U (1 exp( (w wm )/ws ))for w ≥ wm , (2.3) and ϕ(w )=0otherwise,where ws isatypicaldistancebelowwhichthe modificationofthevelocityissignificant(seeFig.2.1),and U isthedesired speed.Typicalvaluesare U =1.25ms 1 , wm =0.3m, ws =0.9m(see Section9.4andFig.9.2therein).

Remark2.1. Followingtheliterature,weadopttheterm“Follow-theLeader”proposedinArgall etal. (2002)todesignatethisapproach.Inthe

Fig.2.1.Speedasafunctionofthedistance.

presentcontext,thetermissomewhat improper,sincethemodelisrather basedonindividualswhohaveapersonalobjective(thatismovingahead atvelocity U ),andwhodecidetoreducetheirspeedtokeepareasonabledistancefromthepersoninfrontofthem,inordertoavoidcollision. Fromthisstandpointeachagentisnotreally led bytheagentahead,but rather disturbed or inhibited.

Underreasonableassumptionsonthebehaviorfunction ϕ,itcanbe provedthattheproblemiswell-posedinthelinearandperiodicsituations, whichmeansthat,forgiveninitialconditions(andprescribedvelocityof N +1inthelinearcase),existenceofa uniquesolutioncanbeestablished. Tobemoreprecise,ifthederivativeof ϕ at wm isnotinfinite,alldistances remainpositiveforalltimes(noaccident).Weputoffuntiltheendofthis sectionarigorousaccountofthoseproperties.

Inordertoinvestigatethestationarystatesofthesystem,itisconvenienttoexpressitintermsofdistances(whichcanbeexpectedtobeconstant,whereaspositionsalwaysmoveforward).Weintroduce wi = xi+1 xi and,sinceweshallbeinterestedinthesituationwherethevelocityof N +1 isconstantandequalto Ueq ,wedefine wN +1 as weq .Thesystemnowwrites

Equilibriumpoint,andstability

Let ϕ bedefinedby(2.3).Inthelinearsetting,weprescribethatthespeed of N +1isaconstant,equalto Ueq ∈]0,U [.Thesystemthenadmitsan

equilibriumpointintermsofdistances:allindividualsmoveatthesame speed Ueq ,andalldistancesareequalto weq ,whichistheonlysolutionto ϕ(weq )= Ueq .Itcanbenumericallycheckedthatiftheinitial conditionisperturbed,theconfigurationreturnstotheequilibrium.This expressestheasymptoticstabilityofthesystem,whichisprovenlocallyby Proposition2.5.Actually,whatevertheinitialcondition,itcanbechecked (andproved,seeCorollary2.1)thatthesystemsystematicallyreturnsto thisequilibriumpoint.

Forcedstopandgoandupstreampropagation

Thespontaneousupstreampropagationcanbeobservedinthefollowing situation.Weconsideragainthelinearsetting,andabehaviorfunction definedby(2.3).Startingfromastationaryevolution(uniformspeed Ueq , equaldistances weq ),wesupposethat N +1suddenlystops,staysstillfor awhile,andthensuddenlyresumemotionatspeed Ueq .Fig.2.2represents theevolutionintimeofthedistances xN +1 xN , xN xN 1 ,.... Numerical solutionisbasedontheexplicitEulerscheme(A.2).Thefigureshowsthat theperturbation(stopandgoforcingoftheheadpedestrian)propagates upstreamwhileundergoingdispersionanddamping(theprofileisstretched andtheamplitudedecreaseswiththedistancetothehead).Wereferto Remark2.6,page24,formoredetailsonthiseffect.

Fig.2.2.Distancevs.time(headdistanceonthetop).

Upstreampropagation:underlyingtransportequation

Weproposehereaninformalexplanationofthetendencyofthesystemto backwardpropagateperturbations,asillustratedbyFig.2.2.Arigorous justificationisproposedinSection2.1.

Weconsidertheequilibriumstateinthecasewheretheheadentity movesatconstantspeed Ueq ∈]0,U [.Allpedestriansmoveatthesame speed,andthecommondistanceissuchthat Ueq = ϕ(weq ).Inother words, W =(weq ,...,weq )isastationarysolutiontothesystem(2.4). Consideraperturbationofthisequilibriumstate:weassumethatdistances areequalto weq + hi ,where hi issmallwithrespectto weq .From(2.4),we havethat

Sincethequantities hi aredefinedatpointswhichareroughlyequidistributedwithaspacestepequalto weq ,wecaninterpretthelastquotient asthediscretizationofthespacederivativeofafunction h(x,t),whichleads (byanti-discretization)totheequation

Itisatransportequationwithconstantvelocity c,with c = weq ϕ (weq ), forwhichsolutionsaretriviallydefinedas h(x,t)= H (x + ct),where H is anysmoothfunction.Thus,itexpressesatransportphenomenonalongthe arrayofindividuals,intheupstreamdirection.Thevelocity c isestimated withinareferentialwhichgloballymovesforwardatspeed ϕ(weq ).Note thatweshallhaveeffectivebackwardpropagationinthefixedreferential wheneverthespeedofthewaveislargerthanthephysicalspeedofentities, i.e.whenever

Inthecasewhereweneglectedthesizeofentities(i.e. wm =0)thefunction ϕ (definedforinstanceby(2.3),with wm =0)istypicallyconcaveand vanishesat0.Thus,anychordintersectsthegraphatauniquepoint, andtheslopeofthegraphissmallerthantheslopeofthechord,i.e. ϕ (weq ) < ϕ(weq ) weq .Inthissituation,thesignaldoesnotgofastenough toactuallymoveupstreaminthefixedreferential.Conversely,ifthesizeof

entitiesisaccountedfor(seeFig.2.1,with wm > 0),wehavetworegimes foreachchord.Thefirstonecorrespondstoalargedensity(smalldistances)andasmallphysicalvelocity,sothatwavepropagatesfasterthan thevelocityofentities,whichresultsinaneffectivebackwardpropagation ofinformationinthefixedreferential.Inanalogywithhydraulicphenomena,thisregimecanbecalled subcritical.Theotherregime(supercritical) correspondstoasmalldensityandalargephysicalvelocity.Notethat, foragivenchord,bothregimecorrespondstothesamepedestrianflow rate ϕ(weq )/weq (thatisthevelocitytimesthedensity).Fromamodelingstandpoint,thesupercriticalregimeischaracterizedbythefactthat perturbationsdonotmigrateupstream,whichmakesiteasytomodela portionofalargersystem.Indeed,ifoneconsidersalargecollectionof individualsmovingrightwardonastraightline,andoneaimsatmodeling thesystemrestrictedtoafixedspaceinterval,itissufficienttoprescribe conditionsattheleftendoftheinterval,wherepeopleenterthechosen window.Inthesubcriticalregime,sinceinformationislikelytoenterthe domainthroughtherightend,suchanapproachisinapplicable,andthe problemofmodelingafixedportionofthepathismuchmoredifficult.

FTLmodel:mathematicalissues

WeaddressinthissectionmathematicalissuepertainingtoModel2.1,i.e. thesystemofdifferentialequations

withprescribedinitialconditions

, and˙xN +1 (t)= V (t)prescribed(or,intheperiodiccase N +1 ≡ 1).

Well-posedness

Letusstartbya“lazy”existencetheorem,lazyinthesensethatwefirst disregardtheissueofcontactsbyextending ϕ on] −∞, 0[(whichdoes notmakemuchsensefromamodelingstandpoint).Thecrucialquestionof actualcontactisputoffuntilthenextproposition.

Proposition2.2. Weconsider N +1 individuals, initiallylocatedat x

Another random document with no related content on Scribd:

The Project Gutenberg

eBook

of Spenser's Faerie Queene, Vol. 2 (of 3)

This ebook is for the use of anyone anywhere in the United States and most other parts of the world at no cost and with almost no restrictions whatsoever. You may copy it, give it away or re-use it under the terms of the Project Gutenberg License included with this ebook or online at www.gutenberg.org. If you are not located in the United States, you will have to check the laws of the country where you are located before using this eBook.

Title: Spenser's Faerie Queene, Vol. 2 (of 3) Books IV-VII

Author: Edmund Spenser

Editor: J. C. Smith

Release date: January 12, 2024 [eBook #72698]

Language: English

Original publication: Oxford: Clarendon Press, 1909

Credits: Tim Lindell and the Online Distributed Proofreading Team at https://www.pgdp.net (This file was produced from images generously made available by The Internet Archive/American Libraries.) *** START OF THE PROJECT GUTENBERG EBOOK SPENSER'S FAERIE QUEENE, VOL. 2 (OF 3) ***

Transcriber’s note

In the HTML version the stanza numbers (indicated by Roman numerals) are in the left margin. In the EPUB3, EPUB, and Kindle versions, they are located on the right side after the first line ends.

The previous volume is available as Project Gutenberg ebook #70717

THE POETICAL WORKS OF EDMUND SPENSER

IN THREE VOLUMES

VOLUME III

SPENSER’S FAERIE QUEENE

EDITED BY

J. C. SMITH

VOLUME II: BOOKS IV-VII

OXFORD

AT THE CLARENDON PRESS

Oxford University Press, Amen House, London E.C.4

GLASGOW NEW

YORK TORONTO MELBOURNE WELLINGTON

BOMBAY CALCUTTA MADRAS KARACHI LAHORE DACCA

CAPE TOWN SALISBURY NAIROBI IBADAN ACCRA

KUALA LUMPUR HONG KONG

FIRST PUBLISHED 1909

REPRINTED LITHOGRAPHICALLY IN GREAT BRITAIN AT THE UNIVERSITY PRESS, OXFORD FROM SHEETS OF THE FIRST IMPRESSION 1961, 1964

THE SECOND PART OF THE FAERIE QVEENE.

Containing

T F, F,

S B.

Imprinted at London for VVilliam Ponsonby. 1596.

THE FOVRTH BOOKE OF THE FAERIE QVEENE.

Containing

The Legend of C and T[1],

OR

OF FRIENDSHIP.

he rugged forhead that with graue foresight i Welds[2] kingdomes causes, and affaires of state, My looser rimes (I wote) doth sharply wite, For praising loue, as I haue done of late, And magnifying louers deare debate; By which fraile youth is oft to follie led, Through false allurement of that pleasing baite, That better were in vertues discipled, Then with vaine poemes weeds to haue their fancies fed.

Such ones ill iudge of loue, that cannot loue, ii Ne in their frosen hearts feele kindly flame: For thy they ought not thing vnknowne reproue, Ne naturall affection faultlesse blame,

For fault of few that haue abusd the same. For it of honor and all vertue is The roote, and brings forth glorious flowres of fame, That crowne true louers with immortall blis, The meed of them that loue, and do not liue amisse.

Which who so list looke backe to former ages, iii And call to count the things that then were donne, Shall find, that all the workes of those wise sages, And braue exploits which great Heroes wonne, In loue were either ended or begunne: Witnesse the father of Philosophie, Which to his Critias, shaded oft from sunne, Of loue full manie lessons did apply, The which these Stoicke censours cannot well deny.

To such therefore I do not sing at all, iv But to that sacred Saint my soueraigne Queene, In whose chast[3] breast all bountie naturall, And treasures of true loue enlocked beene, Boue all her sexe that euer yet was seene; To her I sing of loue, that loueth best, And best is lou’d of all aliue I weene:

To her this song most fitly is addrest, The Queene of loue, and Prince of peace from heauen blest.

Which that she may the better deigne to heare, v Do thou dred[4] infant, Venus dearling doue, From her high spirit chase imperious feare, And vse of awfull Maiestie remoue: In sted thereof[5] with drops of melting loue, Deawd with ambrosiall kisses, by thee gotten From thy sweete smyling mother from aboue, Sprinckle her heart, and haughtie courage soften, That she may hearke to loue, and reade this lesson often.

FOOTNOTES:

[1] Title 5 T] Triamond II xxxi l 8 &c

[2] i 2 Wields 1609

[3] iv 3 chaste 1609 passim

[4] v 2 dred] drad 1609

[5] 5 whereof 1609

Cant. I.

Fayre Britomart saues Amoret, Duessa discord breedes

Twixt Scudamour and Blandamour: Their fight and warlike deedes.

Of louers sad calamities of old, i

Full many piteous stories doe remaine, But none more piteous euer was ytold, Then that of Amorets hart-binding chaine, And this of Florimels vnworthie paine: The deare compassion of whose bitter fit My softened heart so sorely doth constraine, That I with teares full oft doe pittie it, And oftentimes doe wish it neuer had bene writ.

For from the time that Scudamour her bought ii In perilous fight, she neuer ioyed day, A perilous fight when he with force her brought From twentie Knights, that did him all assay: Yet fairely well he did them all dismay: And with great glorie both the shield of loue, And eke the Ladie selfe he brought away, Whom hauing wedded as did him behoue,

A new vnknowen mischiefe did from him remoue.

For that same vile Enchauntour Busyran, iii

The very selfe same day that she was wedded, Amidst the bridale feast, whilest euery man

Surcharg’d with wine, were heedlesse and ill hedded, All bent to mirth before the bride was bedded, Brought in that mask of loue which late was showen: And there the Ladie ill of friends bestedded, By way of sport, as oft in maskes is knowen, Conueyed quite away to liuing wight vnknowen.

Seuen moneths he so her kept in bitter smart, iv Because his sinfull lust she would not serue, Vntill such time as noble Britomart Released her, that else was like to sterue, Through cruell knife that her deare heart did kerue. And now she is with her vpon the way, Marching in louely wise, that could deserue No spot of blame, though spite did oft assay To blot her with dishonor of so faire a pray.

Yet should it be a pleasant tale, to tell v The diuerse vsage and demeanure daint, That each to other made, as oft befell.

For Amoret right fearefull was and faint, Lest she with blame her honor should attaint, That euerie word did tremble as she spake, And euerie looke was coy, and wondrous quaint, And euerie limbe that touched her did quake: Yet could she not but curteous countenance to her make.

For well she wist, as true it was indeed, vi That her liues Lord and patrone of her health

Right well deserued as his duefull meed, Her loue, her seruice, and her vtmost wealth. All is his iustly, that all freely dealth: Nathlesse her honor dearer then her life,

She sought to saue, as thing reseru’d from stealth; Die had she leuer with Enchanters knife, Then to be false in loue, profest a virgine wife.

Thereto her feare was made so much the greater vii Through fine abusion of that Briton mayd: Who for to hide her fained sex the better, And maske her wounded mind, both did and sayd Full many things so doubtfull to be wayd, That well she wist not what by them to gesse[6] , For other whiles to her she purpos made Of loue, and otherwhiles of lustfulnesse, That much she feard his mind would grow to some excesse.

His will she feard; for him she surely thought viii To be a man, such as indeed he seemed, And much the more, by that he lately wrought, When her from deadly thraldome he redeemed, For which no seruice she too much esteemed, Yet dread of shame, and doubt of fowle dishonor Made her not yeeld so much, as due she deemed. Yet Britomart attended duly on her, As well became a knight, and did to her all honor.

It so befell one euening, that they came ix Vnto a Castell, lodged there to bee, Where many a knight, and many a louely Dame Was then assembled, deeds of armes to see: Amongst all which was none more faire then shee, That many of them mou’d to eye her sore. The custome of that place was such, that hee Which had no loue nor lemman there in store, Should either winne him one, or lye without the dore.

Amongst the rest there was a iolly knight, x Who being asked for his loue, auow’d That fairest Amoret was his by right,

And offred that to iustifie alowd. The warlike virgine seeing his so prowd And boastfull chalenge, wexed inlie wroth, But for the present did her anger shrowd; And sayd, her loue to lose she was full loth, But either he should neither of them haue, or both.

So foorth they went, and both together giusted; xi But that same younker soone was ouerthrowne, And made repent, that he had rashly lusted For thing vnlawfull, that was not his owne: Yet since[7] he seemed valiant, though vnknowne, She that no lesse was courteous then[8] stout, Cast how to salue, that both the custome showne Were kept, and yet that Knight not locked out, That seem’d full hard t’accord two things so far in dout.

The Seneschall was cal’d to deeme the right, xii Whom she requir’d, that first fayre Amoret Might be to her allow’d, as to a Knight, That did her win and free from chalenge set: Which straight to her was yeelded without let. Then since that strange Knights loue from him was quitted, She claim’d that to her selfe, as Ladies det, He as a Knight might iustly be admitted; So none should be out shut, sith all of loues were fitted.

With that her glistring helmet she vnlaced; xiii Which doft, her golden lockes, that were vp bound Still in a knot, vnto her heeles downe traced, And like a silken veile in compasse round About her backe and all her bodie wound: Like as the shining skie in summers night, What time the dayes with scorching heat abound, Is creasted all with lines of firie light, That it prodigious seemes in common peoples sight.

Such when those Knights and Ladies all about xiv

Beheld her, all were with amazement smit, And euery one gan grow in secret dout

Of this and that, according to each wit: Some thought that some enchantment faygned it; Some, that Bellona in that warlike wise

To them appear’d, with shield and armour fit; Some, that it was a maske of strange disguise: So diuersely each one did sundrie doubts deuise.

But that young Knight, which through her gentle deed xv Was to that goodly fellowship restor’d, Ten thousand thankes did yeeld her for her meed, And doubly ouercommen, her ador’d: So did they all their former strife accord; And eke fayre Amoret now freed from feare, More franke affection did to her afford, And to her bed, which she was wont forbeare, Now freely drew, and found right safe assurance theare.

Where all that night they of their loues did treat, xvi

And hard aduentures twixt themselues alone, That each the other gan with passion great, And griefull[9] pittie priuately bemone.

The morow next so soone as Titan shone, They both vprose, and to their waies them dight: Long wandred they, yet neuer met with none[10] , That to their willes could them direct aright, Or to them tydings tell, that mote their harts delight.

Lo thus they rode, till at the last they spide xvii

Two armed Knights, that toward them did pace, And ech of them had ryding by his side

A Ladie, seeming in so farre a space, But Ladies none they were, albee in face And outward shew faire semblance they did beare; For vnder maske of beautie and good grace,

Turn static files into dynamic content formats.

Create a flipbook
Issuu converts static files into: digital portfolios, online yearbooks, online catalogs, digital photo albums and more. Sign up and create your flipbook.