简介:Themodulationandcontrolofgecko'sfootmovementswerestudiedelectrophysiologicallyinordertodesignthemotorcontrolsystemofagecko-mimicrobot.Inthisstudy(1)theanatomyoftheperipheralnervescontrollingthegecko'sfootmovementswasdetermined;(2)therelationshipbetweenthelimbnervesofthegeckoanditsfootmotorpatternswasstudied;(3)theafferentimpulsesofthenervesevokedbyrubbingthegecko'stoesandpalmwererecorded;(4)copyingthenaturalpatternsofmovementofthegecko'sfoot(abduction,adduction,flexion,andrevolution)anditslimbnervemodulationandcontrolmechanism,thenerveswerestimulatedundercomputercontrol,andtheresultsrecordedbyCCD.Resultssuggestthatgecko'sfootmovementscanbesuccessfullycontrolledbyartificialelectricalsignals.
简介:AflexiblehingelesscontrolsurfacemodelwasproposedformotioncontrolofUnderwaterVehicles(UVs),whichisinspiredbytheflexiblebendingcontrolsurfacesofunderwatercreatures,suchasfishandsquid.ComputationalFluidDynamics(CFD)simulationdemonstratesthat,incomparisonwiththehingedorrigidcontrolsurface,theproposedflexiblebendingcontrolsurfacecansuppresstheflowseparationsoastoimprovetheturningperformance.Aprototypeoftheflexiblecontrolsurfacewasfabricated,inwhichShapeMemoryAlloy(SMA)wireswereselectedastheactuators.Theelasticenergystorageandexchangemechanismwasincorporatedintotheactuationofthecontrolsurfacetoimprovetheefficiency.ThermalanalysisofSMAwireswasperformedtofindproperactuatingcondition.Open-loopbendingexperimentswerecarriedout.Theresultsshowthattheproposedcontrolsurfacecanachievethemaximumbendingangleof104°.Moreover,thepowerandenergyconsumptionunderdifferentpulseconditionswerecompared.
简介:Asystemisdescribedherethatcannoninvasivelycontrolthenavigationoffreelybehavingratviaultrasonic,epidermalandLEDphoticstimulatorsontheback.Thesystemreceivescommandsfromaremotehostcomputertodeliverspecifiedelectricalstimulationstothehearing,painandvisualsensesoftheratrespectively.Theresultsdemonstratethatthethreestimuliworkingroupsfortheratnavigation.Wecancontroltherattoproceedandmakerightandleftturnswithgreatefficiency.Thisexperimentverifiedthattheratwasabletoreachasettingdestinationinthewayofcablewiththehelpofapersonthroughtheappropriatecoordinationofthethreestimulators.Thetelemetryvideocameramountedontheheadoftheratalsoachieveddistantimageacquisitionandhelpedtoadjustitsnavigationpathoveradistanceof300m.Inaword,thenon-invasivemotioncontrolnavigationsystemisagood,stableandreliablebio-robot.
简介:Aroboticfish,BASEMACK1,isdesignedandfabricatedbymimickingtheshapeofalivemackerel.ThreeDCservo-motorsareseriallylinkedtogetherandactuatedtomimicthemackerel'sCarangiformmotion.Hydrodynamiccharac-teristicsofafish-mimetictestmodelareexperimentallyidentifiedandutilizedinordertonumericallysimulatefishswimming.Thediscretesetofkinematicanddynamicparametersareobtainedbyconsideringrequiredhorizontalandlateralforcesandminimumenergyconsumption.Usingtheoptimizedparameterset,optimalcontroloftherobotisstudied.
简介:OptimalformationreconfigurationcontrolofmultipleUninhabitedCombatAirVehicles(UCAVs)isacomplicatedglobaloptimumproblem.ParticleSwarmOptimization(PSO)isapopulationbasedstochasticoptimizationtechniqueinspiredbysocialbehaviourofbirdflockingorfishschooling.PSOcanachievebetterresultsinafaster,cheaperwaycomparedwithotherbio-inspiredcomputationalmethods,andtherearefewparameterstoadjustinPSO.Inthispaper,weproposeanimprovedPSOmodelforsolvingtheoptimalformationreconfigurationcontrolproblemformultipleUCAVs.Firstly,theControlParameteri-zationandTimeDiscretization(CPTD)methodisdesignedindetail.Then,themutationstrategyandaspecialmutation-escapeoperatorareadoptedintheimprovedPSOmodeltomakeparticlesexplorethesearchspacemoreefficiently.Theproposedstrategycanproducealargespeedvaluedynamicallyaccordingtothevariationofthespeed,whichmakesthealgorithmexplorethelocalandglobalminimathoroughlyatthesametime.SeriesexperimentalresultsdemonstratethefeasibilityandeffectivenessoftheproposedmethodinsolvingtheoptimalformationreconfigurationcontrolproblemformultipleUCAVs.
简介:TheproblemofflappingmotioncontrolofMicroAirVehicles(MAVs)withflappingwingswasstudiedinthispaper.Basedupontheknowledgeofskeletalandmuscularcomponentsofhummingbird,adynamicmodelforflappingwingwasdeveloped.AcontrolschemeinspiredbyhumanmemoryandlearningconceptwasconstructedforwingmotioncontrolofMAVs.Thesalientfeatureoftheproposedcontrolliesinitscapabilitiestoimprovethecontrolperformancebylearningfromexperienceandobservationonitscurrentandpastbehaviors,withouttheneedforsystemdynamicinformation.Furthermore,theoverallcontrolschemehasafairlysimplestructureanddemandslittleonlinecomputations,makingitattractiveforreal-timeimplementationonMAVs.Boththeoreticalanalysisandcomputersimulationconfirmsitseffectiveness.