Unit24IntelligentRobots分解.docx
《Unit24IntelligentRobots分解.docx》由会员分享,可在线阅读,更多相关《Unit24IntelligentRobots分解.docx(10页珍藏版)》请在冰豆网上搜索。
![Unit24IntelligentRobots分解.docx](https://file1.bdocx.com/fileroot1/2023-1/23/5c7f223c-4b1c-46e2-bafd-a461a49c8dec/5c7f223c-4b1c-46e2-bafd-a461a49c8dec1.gif)
Unit24IntelligentRobots分解
PARTⅣAUTOMATIONNEWTECHNOLOGY
Unit24IntelligentRobots
KeyWordsandTerms
1.industrialrobot工业机器人
2.weldv.焊接,使密接
3.assemblev.集合,装配
4.repetitiveadj.重复的,反复性的
5.texturen.质地
6.intelligentrobot智能机器人
7.tactileadj.触觉的,有触觉的
8.simulatev.模拟,模仿
9.bestowv.给予,安放
10.orientationn.方位,定位
11.seamv.缝合,焊合
12.illuminationn.照明,阐明,启发
13.cameran.照相机,摄影机
14.lensn.透镜,镜头
15.pixeln.像素
16.charge-coupleddevices电荷耦合器件
17.randomassessmemory随机存储
18.tactileadj.触觉的,有触觉的
19.wristn.手腕,腕关节
20.straingauge应变片
21.piezoelectric压电的
22.fragileadj.易碎的,脆的
23.navigatev.航行,导航
24.binn.容器,箱子
25.simplyadv.简单地、简明地、朴素地、坦白地、直率地、仅仅、只不过、简直、完全
26.possessvt. 拥有, 持有, 支配
NotationsandTranslation
Industrial[in'dʌstriəl]robots['rəubɔt, 'rəubət]havebeendevelopedtoperformawidevarietyofmanufacturingtasks,suchasmachineloading/unloading,welding,painting,andassembling.However,mostindustrialrobotshaveverylimitedability.Forexample,iftheassemblypartsarenotpresentedtotherobotinaprecise,repetitivefashion,therobotsimplycannotperformitsjobs,becauseitcannotfindthepartandpickitup.Ifaninvadingobjecthappenstoentertheworkspaceofarobot,therobotmaycollidewithit,resultingindamagetoboth,becausetherobotcannotseeitasanobstacle.Inotherwords,existingrobotsarefarfrombeingintelligent,andtheycannothear,see,andfeel(force,shape,texture,temperature,weight,distance,andspeedofobjectsonitsgraspornearby).
换句话说,现有的机器人远还谈不上具有智能,且它们没有听力和视觉,对其手中或附近的物体的力量、形状、质地、温度、重量、距离及速度等还缺乏知觉。
换句话说,现在的机器人还远远不具备智能,而且它们既不能“听”也不能“看”;同时,对于在“手”中或者附近的物体,机器人也缺乏力、形状、纹理、温度、重量、距离和速度等的感知能力。
∙Anentirelynewphaseinrobotic[rəu'bɔtik]
加载中...
applicationshasbeenopenedwiththedevelopmentof“intelligentrobots”.Anintelligentrobotisbasicallyonethatmustbecapableofsensingitssurroundingsandpossess[pə'zes]intelligentenoughtorespondtoachangingenvironmentinmuchthesamewayaswedo.Suchabilityrequiresthedirectapplicationofsensoryperceptionandartificialintelligence.Muchofresearchinroboticshasbeenandisstillconcernedwithhowtoequiprobotswithvisualsensors-eyesandtactile['tæktail, 'tæktil]sensors-the“fingers”.Artificialintelligencewillenabletherobottorespondtoandadapttochangesinitstasksandinitsenvironment,andtoreasonandmakedecisionsinreactiontothosechanges.
智能机器人本质上是具有感知能力并具有足够的智能,在环境改变时做出和人类响应类似的响应。
Inthelightoftheserequirements,anintelligentrobotmaybedescribedasonethatisequippedwithvariouskindsofsensors,possessessophisticated[sə'fistikeitid, sə'fisti.keitid]signalprocessingcapabilitiesandintelligenceformakingdecisions,inadditiontogeneralmobility.
Inthelightof鉴于、由于
Mobilityn.机动性
鉴于以上要求,智能机器人可以形容为,具有多种传感器,具有处理复杂信号的能力且具有做出决策的能力再加上一般的机动性。
Oneimportantcapabilitythathumansdemonstrate['demənstreit, 'demən.streit]whenperformingintelligenttasksisreceivingsensoryinformationfromtheirsurroundings.Humansensesprovidearichvarietyofsensoryinputsthatarecriticaltomanyoftheactivitiesperformed.Muchefforthasbeenmadetosimulatesimilarsensoryabilitiesforintelligentrobots,amongthem,visionisthemostimportantsenseasitisestimatedthatupto80%ofsensoryinformationisreceivedbyvision.Visioncanbebestowed[bi'stəuəd]onroboticsystemsbyusingimagingsensorsinvariousways.Forimprovingaccuracyofperformances,itcanhelppreciselyadjusttherobotbandbymeansofopticalfeedbackcontrolusingvisualsensors.Determiningthelocation,orientation,andrecognitionofthepartstobepickedupisanotherimportantapplication.Forexample,visionisnecessarytoguideaseam-weldingrobottoperformmatingtwopartstogetherandweldthem.Regardlessoftheapplication,thevisionsystemmustcontainanillumination[i.lju:
mi'neiʃən, i.lju:
mə'neiʃən]source,animagery['imidʒəri]sensor,animagerydigitizerandasystemcontrolcomputer.
为了提高工作的精度,可以通过使用视觉传感器,用光学反馈的方式来调节机器手。
识别将要拾起的零件并确定其位置和方向是机器人的另一重要应用。
Ifambientlightingservesanilluminationsource,theimagingprocessispassive.Thistypeofimagingiscommoninmilitaryapplicationssincethepositionoftheviewerisbeyondthecontroloftheviewer.Butinindustrialapplications,controlledilluminationoractiveimagingcanbeusedasfreelyaspossible.
如果周围的光作为照明源,则成像过程为被动的。
这种成像通常用于军事,因为观察者很难控制自己的位置。
但是在工业应用中,受控的照明或称主动成像可以尽可能自由地使用。
Theimagerysensorofarobotvisionsystemisdefinedasanelectroopticaldevicethatconvertsanopticalimage['imidʒ]toavideosignal.TheimagesensorisusuallyeitheraTV-cameraorasolid-statesensorydevice,forinstance,charge-coupleddevices(CCD).Thelatterdeviceoffersgreatersensitivity,longenduranceandlightweight,andisthuswelcomewhencomparedwiththeTV-camera.
机器人视觉系统的图像传感器为将光学图像转换为视频信号的光电设备。
这种图像传感器通常为电视摄像机或者固体传感器件,例如,电荷耦合器件(CCD)。
后者更为灵敏、耐久而且重量轻,因此比电视摄像机更受欢迎。
Thecamerasystemcontainsnotonlythecameradetectorbutalso,andveryimportantly,alenssystem.Thelensdetermined[di'tə:
mind]thefieldofview,thedepthoffocus,andotheropticalfactorsthatdirectlyaffectthequalityoftheimagedetectedbythecamera.Noveltechniques,suchasafish-eyedlensusedtoobtaina360-degreefieldofviewwithouttheneedoffocusadjustment,haverecentlybeeninvestigatedandprovedveryusefulinmobilerobots.
摄像系统不仅仅包含一个照相机,而更重要的,有一个透镜系统。
该透镜确定视野、景深和其他直接影响图像质量的光学参量。
对于鱼眼透镜,可以无须调焦就得到360度的视野。
该新技术,已经证明对于移动机器人非常有用。
EitherTV-camerasorCCDsproduceanimagebygeneratingananaloguevalueoneverypixel,proportionaltoitslightintensity.Toenableadigitalcomputertoworkwithanalogue,ananalogue-to-digital(A/D)converterisneededtotransferanalogueintodigitaldatathenstoredintherandomassessmemory(RAM)installedinthecomputer.Thecomputeranalyzesthedataandextractssuchimageryinformationasedges,regions,boundaries,colorsandtexturesoftheobjectsintheimage.Finally,thecomputerinterpretsorunderstandswhattheimagerepresentsintermsofknowledgeaboutthescene[si:
n]andgivestherobotasymbolicdescriptionofitsenvironment.
无论是电视摄像机还是CCD器件都是在每个象素产生一个与光的强度成比例的模拟量。
为了使数字计算机能够工作,必须要有一个模数转换器将模拟信号转换为数字信号并存在计算机RAM中。
计算机分析数据,并提取出如物体的,边、区域、边界、颜色和纹理等信息。
最后计算机用场景知识解释并理解图像是如何表示的,并向机器人描述其周围的环境。
Nexttovisioninimportanceistactilesensingortouching.Imaginetheblindcandodelicatejobsrelyingonhis/hersensitivetactile!
Ablindrobotcanbeextremelyeffectiveinperforminganassemblytaskusingonlyasenseoftouch.Touchisofparticularimportanceforprovidingfeedbacknecessarytogripdelicateobjectsfirmlywithoutcausingdamagetothem.
仅次于视觉重要性的是触觉。
设想一个盲人可以凭借起敏锐的触觉完成精细的工作。
一个无视觉传感器的机器人可以凭借其触觉十分有效地完成安装工作。
为了牢固地抓起一个精巧的物体而有不对其造成损害,提供反馈触觉非常重要。
Tosimulatetactileinhumanhands,acompletetactile-sensingsystemmustperformthreefundamentalsensingoperations:
(1)jointforcesensingwhichsensetheforceappliedtotherobot’shandwristandarmjoints;
(2)touchsensingwhichsensesthepressureappliedtovariouspointsonthehand’ssurfaceorthegripper’ssurface;(3)slipsensingwhichsensedanymovementoftheobjectwhileitisbeinggrasped.
Thejointforcesareusuallysensedusingvariousstraingaugesarrangedintherobotwristassembly.Astraingaugeisaforce-sensingelementwhoseresistancechangeinproportiontotheamountofforceappliedtotheelement.
Thesimplestapplicationoftouchsensorsisagripperthatisequippedwithanarrayofminiaturemicroswitches.Thistypeofsensorcanonlydeterminethepresenceorabsenceofanobjectataparticularpointoranarrayofpointsoftherobothand.Amoreadvancedtypeoftouchsensorusesarraysofpressure-sensitivepiezoelectric[paɪˌiːzəʊɪ'lektrɪk]material(conductiverubber,conductivefoam,etc.).Thismaterialconductselectricalcurrentwhenstressed.Thisarrangementallowsthesensortoperceivechangesinforceandpressurewithintherobothand.Thesizeofmatricesoftactilesensorsrangesfrom8×8to80×802-Darrays.Sincetheforceateachpointcanbedetermined,theforcesonitssurfacecanbemappedandtheshapesofobjectsgraspedintherobothandcanbedeterminedrespectively.TheforcedatacanbeusedtodisplayonaTV-likescreentheshapeoftheobjectandthedistributionofforceonitssurface.
最简单的接触传感器应用是一个安装了小型的微型开关阵列的爪子。
这种类型的传感器仅仅能够检测一个物体是否在机器人的手上的某点或者点阵列上。
一种更先进的接触传感器是使用压感型的压电阵列材料(导电橡胶、导电泡沫等)。
这种材料在受力时产生电流。
这种传感器可以感知机器人手上的受力或压力。
这种传感器的矩阵尺寸从8╳8到80╳80不等。
因为每一点的力都可以检测出,受力的曲面可以绘出,握在机器人手上的物体的形状可以相应地确定。
受力的数据可以显示在屏幕上以显示物体的形状以及物体表面力的分布。
Slipsensingisrequiredforarobottocreatetheoptimumamountofgraspingforceappliedtoadelicate,fragileobject.Thisabilitypreventsdamagetotheobjectandallowstheobjecttobepickedupwithoutthedangerofbeingdropped.Themethodusedtodetectslipisbasedonsensinganyvibrationoranymovementbetweentheobjectandthegripper,nomatterhowsubtleitmaybe.Thegrippingforceisincreasedstepbystepuntiltheobjecthasbeenfirmlygraspedandnomoreslipoccurs.
Theintegrationoftactilesensingandvisionsensingcandramaticallyenhanceadaptiveroboticassemblytask.Anexampleofthistypeofsensorswouldbeavisionsensorusedtolocateandidentifyobjectsandpositionoftherobotitself,combinedwithatactilesensorusedtodetectthedistributionofforceandpressure,determinetorque,weight,centerofmassandcomplianceofthematerialithandles.
Thishand-eyecoordinationforgeneral-purposemanipulationwouldbeextremelypowerfulintheindustrialworld.
Anotherimportantsensingforarobotisrange,ordepthinformation.Therangedataarenecessaryf