資源描述：
《《acceleration-levelrepetitivemotionplanningofredundantplanarrobotssolvedbyasimplifiedlvi-basedprimal-dualneuraln》由會(huì)員上傳分享，免費(fèi)在線閱讀，更多相關(guān)內(nèi)容在行業(yè)資料-天天文庫(kù)。
1、RoboticsandComputer-IntegratedManufacturing29(2013)328–343ContentslistsavailableatSciVerseScienceDirectRoboticsandComputer-IntegratedManufacturingjournalhomepage:www.elsevier.com/locate/rcimAcceleration-levelrepetitivemotionplanningofredundantplanarrob
2、otssolvedbyasimpli?edLVI-basedprimal-dualneuralnetworka,b,naaaaYunongZhang,HuarongWu,ZhijunZhang,LinXiao,DongshengGuoaSchoolofInformationScienceandTechnology,SunYat-senUniversity,Guangzhou510006,ChinabResearchInstituteofSunYat-senUniversityinShenzhen,S
3、henzhen518057,ChinaarticleinfoabstractArticlehistory:Inthispaper,weproposeanovelrepetitivemotionplanning(RMP)schemeatthejoint-accelerationReceived27July2011level(termed,theacceleration-levelRMPscheme,theARMPscheme),whichincorporatesjoint-angleReceivedi
4、nrevisedformlimits,joint-velocitylimitsandjoint-accelerationlimits.Todothis,Zhangetal’sneural-dynamic4September2012methodisemployedtoderiveanddesignsuchanARMPscheme.SuchaschemeisthenreformulatedasAccepted4September2012aquadraticprogram(QP).TosolvethisQ
5、Pproblemonline,asimpli?edlinear-variational-inequalityAvailableonline18October2012basedprimal-dualneuralnetwork(i.e.,S-LVI-PDNN)isdesigned.Withsimplepiecewise-linearKeywords:dynamicsandglobalexponentialconvergencetotheoptimalsolution,suchanS-LVI-PDNNso
6、lverRepetitivemotionplanningcanhandlethestrictlyconvexQPprobleminaninverse-freemanner.Finally,threegiventasks,Zhangetal’sneural-dynamicmethodi.e.,rhombicpath,straight-linepathandsquarepathtrackingtasks,areful?lledbythree-link,four-linkJoint-acceleratio
7、nleveland?ve-linkplanarrobotarms,respectively.Computer-simulationandphysicalexperimentresultsSimpli?edLVI-basedprimal-dualvalidatethephysicalrealizability,ef?cacyandaccuracyoftheARMPschemeandthecorrespondingneuralnetworkMulti-linkplanarrobotarmsS-LVI-P
8、DNNsolver.&2012ElsevierLtd.Allrightsreserved.1.Introductionlimitsandjoint-accelerationlimits)areusuallynotconsidered,astheseinequalityconstraintsarerelativelynoteasytobeKinematicallyredundantrobotarmsarethosehavingmorehandledviathepseud
