本文摘要：很少有人告诉当手臂拿起一个球时神经、臂膀和传感系统之间的交互。

很少有人告诉当手臂拿起一个球时神经、臂膀和传感系统之间的交互。为了仿真这一大自然反应过程，可以通过微处理器、嵌入式控制软件、执行机构和传感器来结构这一系统从而来研究它们之间的简单关系。这也是美国国防高级研究计划署（DARPA）革命性假肢计划所面对的挑战。美国约翰霍普金斯大学应用于物理实验室是领导性的全球团队，还包括政府机构、大学、私有企业，他们的任务是研发世界上最先进设备的假肢，此假肢由神经输出掌控，使配戴者感觉是一个知道手臂一样需要以一定的速度、灵敏度和力去运动。

先进设备的传感对系统技术需要感官物理输出，如压力、力和温度。这个项目中具备里程碑意义的关键部分是虚拟世界综合环境的研发，一个原始的手臂系统的建模环境用于TheMathworks工具和基于模型设计。虚拟世界综合环境具备标准化的架构和定义完备的界面，需要使二十多有所不同领域专家很好地合作。

TheMathworks工具基于模型设计也被用在其他开发阶段，还包括对臂的机械系统展开建模、测试新的神经解码算法和研发与检验控制算法。为DARPA计划研发的两个原型手臂用于了目标肌肉神经系统，这项技术是由芝加哥康复研究院ToddKuiken博士研发的，内容还包括从被手术手臂到未用于的损害处的肌肉区域的残余神经的传输。在临床评估中，第一个原型需要使患者已完成各种功能任务，还包括从口袋里拿一个信用卡。

VirtualIntegrationEnvironmentArchitectureTheVIEarchitectureconsistsoffivemainmodules:Input,SignalAnalysis,Controls,Plant,andPresentation.TheInputmodulecomprisesalltheinputdevicesthatpatientscanusetosignaltheirintent,includingsurfaceelectromyograms(EMGs),corticalandperipheralnerveimplants,implantablemyoelectricsensors(IMESs)andmoreconventionaldigitalandanaloginputsforswitches,joysticks,andothercontrolsourcesusedbyclinicians.TheSignalAnalysismoduleperformssignalprocessingandfiltering.Moreimportant,thismoduleappliespatternrecognitionalgorithmsthatinterpretrawinputsignalstoextracttheusersintentandcommunicatethatintenttotheControlsmodule.IntheControlsmodule,thosecommandsaremappedtomotorsignalsthatcontroltheindividualmotorsthatactuatethelimb,hand,andfingers.ThePlantmoduleconsistsofaphysicalmodelofthelimbsmechanics.ThePresentationmoduleproducesathree-dimensional(3D)renderingofthearmsmovement(Figure1).InterfacingwiththeNervousSystemSimulinkandtheVIEwereessentialtodevelopinganinterfacetothenervoussystemthatallowsnaturalandintuitivecontroloftheprostheticlimbsystem.Researchersrecorddatafromneuraldeviceimplantswhilethesubjectsperformtaskssuchasreachingforaballinthevirtualenvironment.TheVIEmodularinputsystemsreceivethisdata,andMATLABalgorithmsdecodethesubjectsintentbyusingpatternrecognitiontocorrelateneuralactivitywiththesubjectsmovement(Figure2).TheresultsareintegratedbackintotheVIE,whereexperimentscanberuninrealtime.Thesameworkflowhasbeenusedtodevelopinputdevicesofallkinds,someofwhicharealreadybeingtestedbyprostheticlimbusersattheRehabilitationInstituteofChicago.BuildingReal-TimePrototypeControllersTheSignalAnalysisandControlsmodulesoftheVIEformtheheartofthecontrolsystemthatwillultimatelybedeployedintheprostheticarm.AtAPL,wedevelopedthesoftwareforthesemodules.IndividualalgorithmsweredevelopedinMATLABusingtheEmbeddedMATLABsubsetandthenintegratedintoaSimulinkmodelofthesystemasfunctionblocks.Tocreateareal-timeprototypeofthecontrolsystem,wegeneratedcodeforthecompletesystem,includingtheSimulinkandEmbeddedMATLABcomponents,withReal-TimeWorkshop,anddeployedthiscodetoxPCTarget.Thisapproachbroughtmanyadvantages.UsingModel-BasedDesignandSimulink,wemodeledthecompletesystemandsimulatedittooptimizeandverifythedesign.Wewereabletorapidlybuildandtestavirtualprototypesystembeforecommittingtoaspecifichardwareplatform.WithReal-TimeWorkshopEmbeddedCoderwegeneratedtarget-specificcodeforourprocessor.BecausethecodeisgeneratedfromaSimulinksystemmodelthathasbeensafety-testedandverifiedthroughsimulation,thereisnohand-codingstepthatcouldintroduceerrorsorunplannedbehaviors.Asaresult,wehaveahighdegreeofconfidencethattheModularProstheticLimbwillperformasintendedanddesigned.PhysicalModelingandVisualizationToperformclosed-loopsimulationsofourcontrolsystem,wedevelopedaplantmodelrepresentingtheinertialpropertiesofthelimbsystem.WebeganwithCADassembliesoflimbcomponentsdesignedinSolidWorksbyourpartners.WeusedtheCADassembliestoautomaticallygenerateaSimMechanicsmodelofthelimblinkedtoourcontrolsysteminSimulink.Finally,welinkedtheplantmodeltoaJava3DrenderingenginedevelopedattheUniversityofSouthernCaliforniatoshowavirtuallimbmovinginasimulatedenvironment.functionImgZoom(Id)//新的设置图片大小避免撑破表格{varw=$(Id).width;varm=650;if(w。

本文来源：k8凯发官网入口-www.wncyxy.com