DevelopingaVisualizationToolforSpiderWeb-BuildingAlgorithms

模擬蜘蛛結網(wǎng)之演算法設計及視覺化工具開發(fā)

指導教授：尹邦嚴陳怡孜陳瑩哲

沈扇綸

郭怡君1/11/20231OPlab,IM,NTUOutlineIntroductionAbstractMotivationLiteraturereviewArtificialIntelligenceRelatedspiderpapersSpiderweb-buildingalgorithmdesignVisualizationtoolprototypeofthemodelInternalfactorsExternalfactorsConclusionandfuturework1/11/20232OPlab,IM,NTUI.Introduction 1/2AbstractSequentialbehaviorofspiderweb-buildingmodelInternalfactorsofthespiderweb-buildingmodelExternalfactorsofthespiderweb-buildingmodelImplementationofvisualizationtool1/11/20233OPlab,IM,NTUI.Introduction 2/2MotivationFigure1.Themotivationofthemodel1/11/20234OPlab,IM,NTUOutlineIntroductionAbstractMotivationLiteraturereviewArtificialIntelligenceRelatedspiderpapersSpiderweb-buildingalgorithmdesignVisualizationtoolprototypeofthemodelInternalfactorsExternalfactorsConclusionandfuturework1/11/20235OPlab,IM,NTUII.Literaturereview1/3ArtificialIntelligence

SwarmintelligenceGeesemigratine–flyina“V”formationBeesandwag-taildanceAntcolonyoptimization(ACO) Solvediscreteoptimizationproblems(Dorigo,1992)Particleswarmoptimization(PSO)Aoptimizationforrepresentingsociocognitionofhumanandartificialagentstool(KennedyandEberhart,1995)1/11/20236OPlab,IM,NTUII.Literaturereview 2/3RelatedspiderpapersInfluenceofthegenesontheweb(Kirnk,1996)Wind,Temperature,andHumidity(Vollrath,1997)Preysize(Vollrath,1998)Thepatternofthecapturespiral(李蔡彥and林翰儂,2004)1/11/20237OPlab,IM,NTUII.Literaturereview 3/3RelatedspiderartificialintelligenceThepatternofthecapturespiral(李蔡彥and林翰儂,2004)Figure.2Capturespiral patternFigure.3Differentcapturespiralpattern1/11/20238OPlab,IM,NTUOutlineIntroductionAbstractMotivationLiteraturereviewArtificialIntelligenceRelatedspiderpapersSpiderweb-buildingalgorithmdesignVisualizationtoolprototypeofthemodelInternalfactorsExternalfactorsConclusionandfuturework1/11/20239OPlab,IM,NTUIII.Spiderweb-building

algorithmmodeldesign1/8Figure.4Thefactor-analysistreeofweb-building.1/11/202310OPlab,IM,NTUIII.Spiderweb-buildingalgorithmmodeldesign 2/8Figure.5Therule-analysistreeofweb-building.

Auxiliaryspiral(AS)Capturespiral(CS)Figure.6Therealspiderweb.

Influenceofthegenesontheweb(Kirnk,1996)1/11/202311OPlab,IM,NTUIII.Spiderweb-buildingalgorithmmodeldesign 3/8RadiiGene1：Controlthenumberofradii()andbaseangle().

Gene2：Controlthedifferenceoftheangle()fromthenorthtothefirstradii.Gene3：Controlthedifferenceoftheangle()betweentheotherradii.Figure.7Therelationbetween radiiandgene1/11/202312OPlab,IM,NTUIII.Spiderweb-buildingalgorithmmodeldesign 4/8FrameThedistance(HF)betweenhubandframeisthelengthoftheradii(RL)subtractthelengthwhichiscontrolbyGene4(

). HF=RL-Becauseoftheeffectsofthegravity,thedistancebetweenhubanddownwardframeislongerthantheothersdistance.RLHFFigure.8Therelation between frame andgene1/11/202313OPlab,IM,NTUIII.Spiderweb-buildingalgorithmmodeldesign 5/8ASTheeffectsoftheinteractionsbetweenGene5(

)andGene6(

)controltheintersectionpointsoftheASandradii,resultinginaclockwisespiral(AS).Figure.9Therelationbetween ASandgene1/11/202314OPlab,IM,NTUIII.Spiderweb-buildingalgorithmmodeldesign 6/8CS AlongtheAS,aspidercreatesananticlockwisespiral(CS).TheintersectionpointsoftheCSandradiiareeffectedbyGene7(

).Figure.10Therelationbetween CSandgene1/11/202315OPlab,IM,NTUIII.Spiderweb-buildingalgorithmmodeldesign 7/8CSreverse Usingreversetofilltheremainingspaceofweb,whichisuncoveredwithCS.Figure.11Therelationbetweenradii andgene1/11/202316OPlab,IM,NTUIII.Spiderweb-buildingalgorithmmodeldesign 8/8Figure.12Sequentialbehaviorofspiderweb-building.Experiments1/11/202317OPlab,IM,NTUOutlineIntroductionAbstractMotivationLiteraturereviewArtificialIntelligenceRelatedspiderpapersSpiderweb-buildingalgorithmdesignVisualizationtoolprototypeofthemodelInternalfactorsExternalfactorsConclusionandfuturework1/11/202318OPlab,IM,NTUIV.Visualizationtoolprototypeofthemodel 1/23Internalfactors

ExplanationoftherulesExperimentsModelsimulationVerificationExternalfactorsExplanationoftherulesExperimentsModelsimulationVerification1/11/202319OPlab,IM,NTUIV.Spiderweb-buildingalgorithmmodeldesign 2/12InternalfactorsFigure.6Theinternalfactor-analysistreeofweb-building1/11/202320OPlab,IM,NTUIV.Visualizationtoolprototypeofthemodel 3/23InternalfactorsBodysize(χ) LargerspiderswouldbuildlargerwebsWeight(ψ) HeavierspiderswiththesamesizeoftheirbodywouldbuildlargerwebsGlandsilk(ω) Largerglandsilkwiththesamesizeoftheirbodybuildlargerwebsχ,ψ,ωinteraction1/11/202321OPlab,IM,NTUIV.Visualizationtoolprototypeofthemodel 3/23Figure.13(a)minimizeχ,ψ,ωFigure.13(b)Maximizeχ,minimizeψ,ωBodysize(χ)experimentsModelsimulationVerification1/11/202322OPlab,IM,NTUFigure.14(a)minimizeχ,ψ,ωFigure.14(b)Maximizeψ, minimizeχ,ωIV.Visualizationtoolprototypeofthemodel 4/23Weight(ψ)experimentsModelsimulationVerification1/11/202323OPlab,IM,NTUIV.Visualizationtoolprototypeofthemodel 5/23Glandsilk(ω)experimentsModelsimulationVerificationFigure.15(b)Maximizeω, minimizeχ,ψFigure.15(a)minimizeχ,ψ,ω1/11/202324OPlab,IM,NTUIV.Visualizationtoolprototypeofthemodel 6/23Figure.16(b)Maximizeω,χ,ψχ,ψ,ωinteractionexperiments

experimentsModelsimulationVerificationFigure.16(a)minimizeχ,ψ,ω1/11/202325OPlab,IM,NTUIV.Visualizationtoolprototypeofthemodel 1/23Internalfactors

ExplanationoftherulesExperimentsModelsimulationVerificationExternalfactorsExplanationoftherulesExperimentsModelsimulationVerification1/11/202326OPlab,IM,NTUIV.Visualizationtoolprototypeofthemodel 7/23Figure.17Theexternalfactoranalysistreeofweb-buildingExternalfactorsClimateSurroundingenvironmentsPreysize1/11/202327OPlab,IM,NTUIV.Visualizationtoolprototypeofthemodel 8/23ClimateWindAffectmeshAffectthelengthandthenumberofradiusTemperatureAffectmeshAffectthefrequencyofCSreverseHumidityAffectmeshAffecttheviscosityofCS1/11/202328OPlab,IM,NTUIV.Visualizationtoolprototypeofthemodel 9/23Wind()experimentsModelsimulation AdjustingthecontrolvalueofGenetocreatenewmesh.Figure.18Windinfluenceinmesh1/11/202329OPlab,IM,NTUIV.Visualizationtoolprototypeofthemodel 10/23：0m/s：25℃

：45%：1m/s：25℃：45%Verification-wind

Bias:16.33%Bias:17.33%Figure.19(a)WindeffectmeshFigure.19(b)Windeffectmesh1/11/202330OPlab,IM,NTUTemperature()experimentsModelsimulation AdjustingthecontrolvalueofGenetocreatenewmesh.IV.Visualizationtoolprototypeofthemodel 11/23Figure.20Temperatureinfluenceinmesh1/11/202331OPlab,IM,NTUIV.Visualizationtoolprototypeofthemodel 12/23：24℃：0m/s

：55%：12℃：0m/s

：55%Verification-temperature

Figure.21(a)TemperatureeffectmeshFigure.21(b)TemperatureeffectmeshBias:0.60%Bias:1.67%

1/11/202332OPlab,IM,NTUHumidity()experimentsModelsimulation AdjustingthecontrolvalueofGenetocreatenewmesh.IV.Visualizationtoolprototypeofthemodel 13/23Figure.22Humidityinfluenceinmesh1/11/202333OPlab,IM,NTUVerification-humidityIV.Visualizationtoolprototypeofthemodel 14/23

：70%：0m/s

：24℃：20%：0m/s

：24℃Figure.23(a)HumidityeffectmeshFigure.23(b)HumidityeffectmeshBias:0.87%Bias:0.16%

1/11/202334OPlab,IM,NTUSurroundingenvironmentsToaffecttheareaofthewebThegoalofweb-buildingistomaximizeIV.Visualizationtoolprototypeofthemodel 15/231/11/202335OPlab,IM,NTUIV.Visualizationtoolprototypeofthemodel 16/23Figure.24(a)Availablespace relativelysmallSurroundingenvironmentsexperimentsModelsimulationVerificationFigure.24(a)Availablespace relativelybig1/11/202336OPlab,IM,NTUIV.Visualizationtoolprototypeofthemodel 17/23PreysizeAffectmeshApplyingthepastexperiencesforcatchingdifferencesizesofpreytoadjustthemeshofthenextweb-building1/11/202337OPlab,IM,NTUIV.Visualizationtoolprototypeofthemodel 18/23PreysizeexperimentsModelsimulation ThemajorconceptisthatusinganimalslearningtosimulatetherealworldVerification1/11/202338OPlab,IM,NTUIV.Visualizationtoolprototypeofthemodel 19/23Figure.25(a)ThefirstdayFigure.25(b)TheseconddayVerification1/11/202339OPlab,IM,NTUIV.Visualizationtoolprototypeofthemodel 20/23VerificationFigure.25(c)ThethirddayFigure.25(d)Thefourthday1/11/202340OPlab,IM,NTUIV.Visualizationtoolprototypeofthemodel 21/23Summary1/11/202341OPlab,IM,NTUOutlineIntroductionAbstract