KINEMATIC MODELING OF ROBOTIC HAND

Transcription

1 KINEMAI MODELING OF ROBOI HAND Hussan M.Z, Garga R, Assocate professor, Department of Mechanca Engneerng, IIM oege of Engneerng,Greater Noda,(Inda) ABRA hs paper presented a knematc and anass of mut-fngered robotc hand. he paper has an nsght toward the deveopment of the theoretca framework and anatca resuts n ths area of mut fngered robotc hands w be presented net pubcaton. We man addressed research that has estabshed the theoretca framework for knematc anass, smuaton and snthess. Because of the mtatons on space, we have not gven the agorthmc aspects and the appcatons the attenton that the deserve. We hope that ths paper fnd appcaton for further work n the fed of robotcs hand modeng and anass I. INRODUION A robot s an automatc mechanca devce often resembng a human or anma. Modern robots are usua an eectro-mechanca machne guded b a computer program or eectronc crcutr. Robots can be autonomous or sem-autonomous and range from humanods to ndustra robots, coectve programmed swarm robots, and even mcroscopc & nano robots. Robotc hand s the end effector, or robotc hand, can be desgned to perform an desred task such as wedng, grppng, spnnng etc., dependng on the appcaton. he of Robotc hands are cassfed as two fngered robotc hand and mut-fngered robotc hand (MFRH) he Mut-fngered robot hand have aread been desgned and fabrcated to accommodate a varet of tasks such as graspng and manpuaton of objects n the fed of ndustra appcatons. he frst step n reasng a fu functona Mut-fngered robot hand s mathematca modeng. A mut fngered robotc hand mode s studed based on the boogca equvaent of human hand. Fg.. Mode of MFRH 496 P a g e

2 II. LIERAURE REVIEW J. Denavt and R.. Hartenberg [] descrbed an robot knematca b gvng the vaues of four quanttes for each nk. wo descrbe the nk tsef, and two descrbe the nk s connecton to a neghborng nk. In the usua case of a revoute jont, s caed the jont varabe, and the other three quanttes woud be fed nk parameters. For prsmatc jonts, d s the jont varabe, and the other three quanttes are fed nk parameters. he defnton of mechansms b means of these quanttes s a conventon usua caed the Denavt- Hartenberg notaton. John J.rag. [] Presented a technque for conventon and notaton whch can be used to deveop the knematcs and dnamcs equaton. Vctor J. Johnson, Gregor P. tarr [] deveops the knematc and dnamc equatons for one fnger of the three-fngered tanford/jpl robot hand and documents the phsca parameters needed to mpement the equatons.. Parasuraman, and Ler haw Pe [4], the bomechanca anass of human jonts s carred out and the stud s etended to the robot manpuator.. Vaentn Grecu, et. A. [5] he artcuate the fnger conssts of a set of rgd segments connected wth jonts. Each fnger jont anges w be computed b the gven fngertp poston and orentaton. Izzedn Ibrahm Mohamed Abdeazz [6], the focus on the desgn of compact fve fngers rea tme smart gove to emuate the human hand functons that can be used as a prototpe mode for hand rehabtaton sstems for patents sufferng from paraze or contracture. hs work tred to emuate the grp functon and deveop a prototpe mode. Kazuhro Kosuge, Katsuhsa Furuta [7], the have presented a quanttatve measure of controabt of robot arms. hs measure s representatve of the nput- output reaton of a robot arm at a pont, and reated to both knematc and dnamc propertes of the arm ncudng actuators. Wan Fazura Bnt Wan armz, et a [8], the compete dervaton of the mathematca modeng comprsng the knematcs and dnamc of mut fngered robotc hand was carred out b usng Denavt Hartenburg (DH) agorthm[] and Euer Langrange formua [] to enabe subsequent smuaton work. Oho.. and Kajae. R. [9], A Bommetc approach s adopted for the grpper desgn because the stabe grasp can on be acheved wth mutfngered grppers..ramasam and M.R.Arshad [], the work presented n ths paper s to smuate a robotc hand that emuates the shape and performance of a human hand (.e. pam and fngers secton). he graphc desgn s used as a foundaton to fnd the knematcs and dnamc propertes of the robotc hand. E.A. A-Gaaf [], present a nove neura network for deterous hand-graspng nverse knematcs mappng used n force optmzaton. hs s done b consderng the nverse hand Jacoban, n addton to the nteracton between hand fngers and the object. H. Jack, et. a. [], the use of feed forward neura networks to sove the nverse knematcs probem s eamned for three dfferent cases. A cosed knematc nkage s used for mappng nput jont anges to output jont anges... OHOL &. R. KAJALE [], the focused on enhancng the graspng abt wth better sensors backup, whch can enabe the robot to dea wth rea fe stuatons. to desgn grpper b epermentng wth varous desgns for deveopng the unversa deterous graspng. Muhammad E. Abdaah, et a [4], the presents a jont-space torque contro aw that demonstrates both a decouped and sgnfcant faster response than an equvaent tendon-space formuaton. A tenson dstrbuton agorthm s presented here to aocate forces from the jonts to the tendons. he contro aw and tenson dstrbuton agorthm are mpemented on the robotc hand of Robonaut-. Jasper chuurmans et a [5], the dscussed 497 P a g e

3 the optmzaton of the rato between hghest and owest force on the proma phaanges from 4. to.9 whch tend to drve objects out of the grasp when the fnger fees. Guan Ysheng & Zhang Qan [6], the paper nvestgates manpuaton of an object nteractng wth the envronment. Poston manpuaton was accompshed b three knematc agorthms, and force manpuaton was fufed through poston contro. III. FORWARD KINEMAI Fg. 4 shows the mode of a fnger whch has four frames attached to ts jonts. he base frame s referrng the fed jont, whch s ndcated as X, Y, Z. X, Y, Z and X, Y, Z are representng jont and jont respectve, whereas X,Y, Z s representng the fngertp poston. Fg. 4: Mode of One Fnger Forward knematc s used to determne the poston and orentaton of the fngertp reatve to the robot base coordnate sstem. he dervaton of forward knematc equaton s done as foows. Where, = Jont ange of the fnger d = jont dstance of the fnger a = nk ength of the each jont = nk twst ange Fg.5. Denavt Hartenberg (DH) frame [] 498 P a g e

4 499 P a g e he foowng Homogeneous transform equatons are used to determne the transform between base frame X, Y, Z to the fnger tp frame X 4, Y 4, Z 4. Equaton () s the generazed equaton representng the transformaton between the frames - and. d d a () Forward Knematc s used to determne the poston and orentaton of mut fngered robotc hand to determne the poston and orentaton of the robot hand reatve to the robot base (Pam) coordnate sstem. he dervaton of forward knematc equaton based on foowng abe abe : Fnger Parameters d a (MP) (PIP) 4 4 (DIP) ---() ---() ---(4) herefore, we can cacuate (5) -- (6)

5 5 P a g e 4 ---(7) herefore, the forward knematc for the fngers of robot hand are gven b: (8) 4 -- (9) IV. INVERE KINEMAI For nverse knematc, we are gven artesan coordnates,,, and φ. We w be fndng the vaue for,, uch as, () Inverse knematcs has mutpe soutons for a specfc poston and orentaton of the fnger tp. If there s a no souton that means the fnger cannot attan the gven poston and orentaton because t es outsde of the fnger workspace. he estence or nonestence of a knematc souton defnes the workspace of a gven fnger. Let the gven orentaton be n () Fg. 6: Feon Anges of Fngers

6 () Frst of a we compare equaton (6) and () we have for the fngers () (4) quarng both sde and addng equaton (5) and (6) we have to emnate ( ) ( ) ( ) ( ) ( ) ( ) (5) Equaton (5) s n smar form of Pcos Qsn R (6) Let r cos (7) rsn (8).e., arctan ---- (9) r () And b usng above there are two soutons for cos () 5 P a g e

7 5 P a g e, tan a (), arctan () quarng and addng agan equaton (5) and (6) we have emnate ) ( ) ( (4) Aso we can cacuate and, tan A (5),, tan A (6) ( ) V. FINDING he paper [6] eft epore and move functon of the human hand. Aso t can use for modeng of man fngered robotc hand. he paper [7] has not consdered nonnear propertes such as the effect of the oros and centrfuga term. For the desgn of the controer of such an arm, ths nonneart s ver mportant. he resuts of the paper [8] have not pubshed n t now. he presented mechansm [9] posses scope of deveopment. It can grp odd sze and odd shape object, and further t can be deveoped for unversa grpper sstem. he grasp sstem can be desgned to get perfect grasp ponts usng a vson sensors nput can offer a unversa bommetc grpper sstem. he resuts presented n paper [] w be hepfu n utzng the estng graphc software for conductng robotc hand smuaton, and understandng how to reate the knematc and dnamc anass to the actua smuaton procedure.

8 VI. ONLUION hs paper presented aknematc modeng and anass of mut-fngered robotc hand. It s not possbe to justf to a the work n ths area, partcuar throughout the fed and ts deterous manpuaton, epore, move, and to grp objects. We presented about knematc modeng and anass of mut-fngered robotc hand. In addton, we man addressed research that has estabshed the theoretca framework for knematc anass, smuaton and snthess. Because of the mtatons on space, we have not gven the agorthmc aspects and the appcatons the attenton that the deserve. We hope that ths paper fnd appcaton for further work n the fed of robotcs hand modeng and anass. REFERENE. J. Denavt and R.. Hartenberg, "A Knematc Notaton for Lower-Par Mechansms Based on Matrces," Journa of Apped Mechancs, pp. 5, June John J.rag. Introducton to Robotcs. nd ed. Pearson Educaton Internatona.. Vctor J. Johnson, Gregor P. tarr., Knematc and Dnamc Anases of the tanford/ JPL Robot Hand, anda report presented b anda Natona Laboratores, November Parasuraman, and Ler haw Pe., Bo-mechanca Anass of Human Jonts and Etenson of the tud to Robo, Word Academ of c., Engg and ech, Vaentn Grecu, et. A., Anass of Human Arm Jonts and Etenson of the tud to Robot Manpuator, Proceedngs of the Internatona Mut-onference of Engneers and omputer centsts 9 Vo II IME 6. Izzedn Ibrahm Mohamed Abdeazz., A Low-ost mart Gove for Hand Functons Evauaton, Internatona Journa of Robotcs and Automaton (IJRA) Vo., No., March 4, pp. 9~5 7. Kazuhro Kosuge, Katsuhsa Furuta., Knematc And Dnamc Anass Of Robot Arm, IEEE pp-9-44, [8] Wan Fazura Bnt Wan armz, et a, Knematc and Dnamc Modeng of a Mut-Fngered robot Hand, Internatona Journa of Basc & Apped cences Vo: 9 No: page OHOL.. and KAJALE. R., Bommetc Approach for Desgn of Mutfngered Robotc Grpper(MRG) & Its Anass for Effectve Deterous Graspng, 9 Internatona onference on Machne Learnng and omputng IPI vo. (),page--,ngapore..ramasam and M.R.Arshad, Robotc hand smuaton wth knematcs and dnamc anass, IEEE () page E.A. A-Gaaf, Mut-fngered robot hand optma task force dstrbuton Neura nverse knematcs approach, Internatona Journa of Robotcs and Autonomous stems 54 (6) page H. Jack, D. M. A. Lee, R. O. Buchaand W. H.Emaragh, Neura networks and the nverse knematcs probem, Journa of Integent Manufacturng (99) 4, OHOL,. R. KAJALE, muaton of Mutfnger Robotc Grpper for Dnamc Anass of Deterous Graspng, Proceedngs of the Word ongress on Engneerng and omputer cence, October -4, 8, an Francsco, UA. 5 P a g e

9 4. Muhammad E. Abdaah, et a, Apped Jont-pace orque and tffness ontro of endon-drven Fngers, IEEE-RA Internatona onference on Humanod Robots Nashve, N, UA, December 6-8, 5. Jasper chuurmans et a, Grasp force optmzaton n the desgn of an underactuated robotc hand, Proceedngs of the 7 IEEE th Internatona onference on Rehabtaton Robotcs, June -5, Noordwjk, he Netherands pp Guan Ysheng, Zhang Qan, hbrd poston/ force manpuaton of mutfngered hand, hnese Journa Of Aeronautcs, Vo. No., pp-4-5, Ma Yoshhko Nakamura and Katsu Yamane, Dnamcs omputaton of tructure-varng Knematc hans and Its Appcaton to Human Fgures, IEEE ransactons On Robotcs And Automaton, Vo. 6, No., Apr pp K. aha, Dnamcs of era Mutbod stems Usng the Decouped Natura Orthogona ompement Matrces, ransactons of the AME, Vo. 66, pp , December, Wam W. Armstrong and Mark W. Green, he dnamcs of artcuated rgd bodes for purposes of anmaton, he Vsua omputer (985) I: pp--4. E. A. AL-GALLAF, Neurofuzz Inverse Jacoban Mappng for Mut-Fnger Robot Hand ontro, Journa of Integent and Robotc stems 9: pp-7 4, 4.. Rajko omovc, et a, A trateg For Grasp nthess Wth Mutfngered Robot Hands, Proceedng of IEEE onference on Robotcs and Automaton, pp-8-89, Antono Bcch and Vja Kumar, Robotc Graspng and ontact: A Revew, Robotcs and Automaton,. Proceedngs. IRA'. Pp-48-5 IEEE Internatona onference. herr Laberte, et a, Underactuaton n robotc graspng hands, Machne Integence & Robotc ontro, Vo. 4, No., pp-,. 4. Panagots K. Artemads and Kostas J. Krakopouos, EMG-based Poston and Force ontro of a Robot Arm: Appcaton to eeoperaton and Orthoss 5. Okada,.; sucha,. : "On a Versate Fnger stem", Proc. of the 7th Int. mposum Of Industra Robots, Okada,.: "omputer ontro of Mutjoned Fnger stem for Precse Object Handng, IEEE rans. on stems, Man and bernetcs, vo. (), Ma/June Jacobson,..; Wood, J.E.; Knutt, D.F.; Bggers, K.B.: "he Utah/MI Detrous Hand: Work n Progress", Frst Int. onf. on Robotcs Research, MI Press, Jacobson,.. et.a.: "Desgn of the Utah/MI Detrous Hand", Proc. of the IEEE Int. onf. on Robotcs and Automaton, Narasham,.: "Detrous Robotc Hands: Knematcs and ontro", Master thess, Department of Eectrca Engneerng and omputer cence, MI, asbur, K.; Brock, D.; hu,.: "Integrated Language, ensng and ontro for a Robot Hand", Proc. of Int. mp. on Robotcs Research, MI Press, Nguen, V.: "onstructng tabe, Force-osure Grasps", Master hess, Department of Eectrca Engneerng and omputer cence, MI, P a g e

10 . Arbb, M.A.; Ibera,.; Lons, D.: "oordnated ontro Programs for Movements of the Hand enter of stems Neuroscence and Laborator for Perceptua Robotcs, OIN echnca Report 8-5, Massachusetts, 98.. Lons, D.M.: "A smpe et of Grasps for a Detrous Hand", Proc. of the IEEE Int. onf. on Robotcs and Automaton, Ibera,.: "Grasp Pannng for Human Prehenson", Proc. of the Int. Jont onf. on Artfca Integence, Jacobsen, E.K. Inversen, D.F. Knutt, R.. Johnson and K.B. Bggers, Desgn of the Utah/MI Deterous Hand, Proceedngs of the 986 IEEE Internatona onference on Robotcs and Automaton, pp 5-5, P a g e