多连杆冗余移动机械臂建模与运动规划

整理文档很辛苦,赏杯茶钱您下走!

免费阅读已结束,点击下载阅读编辑剩下 ...

阅读已结束,您可以下载文档离线阅读编辑

资源描述

0,,,[1].,:,[2];,[3].,.,.,,23[4,5].1,2,1,2,1,2(1.,200237;2.,310018):,,.,,,.,.,,.:;;;:TP18:A:1001-7119(2010)02-0200-06ModelingandMotionPlanningforMulti-LinkRedundantMobileManipulatorZHANGBotao1,2,LIUShirong1,2,SHIXianpeng1,2(1.ResearchInstituteofAutomation,EastChinaUniversityofScienceandTechnology,Shanghai200237,China;2.ResearchInstituteofAutomation,HangzhouDianziUniversity,Hangzhou310018,China)Abstract:Todealwiththepotentialfieldtrapproblemofartificialpotentialfield(APF)andtheworkspacelimitationofplanarmanipulator,aglobalmotionplanningmethodisproposedforredundantmobilemanipulator.Inthisstudy,repul-sivepotentialfunctionismodifiedbyintroducingasafetyfactorintoAPFtopreventmobilemanipulatorfromgoalun-reachableproblemandtheparametersofpotentialfieldmodelareoptimizedbyquantumgeneticalgorithm(Q-GA).Toimprovethepositioningaccuracy,anewparticleswarmoptimization(PSO)algorithmisappliedtotheinversekinematicsproblemofmobilemanipulator.SimulationresultshowsthattheproposedmethodcaneffectivelyovercomethedefectsofAPFmodelandreducethetimeanderrorwhiletheend-effectorreachesthedesiredposition.Keywords:mobilerobot;artificialpotentialfield;quantumgeneticalgorithm;particleswarmoptimization:2009-09-15:(60675043);(2007C21051);(KYS09150543):(1982-),,,,。E-mail:liushirong@163.com.Vol.26No.2Mar.201026220103BULLETINOFSCIENCEANDTECHNOLOGY21Fig.1Assignmentoflinkframesforthemobilemanipulator[6,7],,,,.,,.,.,Q-GA,,Arm-PSO..,,Q-GA,;,Arm-PSO、.,.、.:iRobotSUGVEarly、PackBot.1,,,,,。,,,,,1。1.11,Ts,Tr,Tg,T0~T516,T6.:(),、、alength,awidth,aheight.z0,yi,i=1,2,…5….,.w1,w2,w3,w4,w5,w6,h,、:sTr=cosφ-sinφ0xrsinφcosφ0yr00100001!#$%%%%%%%%%&,topT0=10000100001awidth+h0001!#$%%%%%%%%%&.:rT0=cosθ10-sinθ1w1cosθ1sinθ10cosθ1w1sinθ0-1000001!#$%%%%%%%%%&,,φ,xr,yrx,y,θ11.:θi=0,i=1,2,…,6.,yi.,.201266i=2Σθi≤π2.,,[-π/2,π/2],ki=2Σθi≤π2,k=2,…,6.,,T0,T0(pgx,pgy,pgz),,:p6x=w1+6i=2Σwicos(ij=2Σθj)p6y=0p6z=6i=2Σwisin(ij=2Σθjjjjjjjjjjjjjjjjjjjjjj).(1),θji.p6x,p6y,p6z.,,,:ki=2Σθi≤π2,k=2,…,6.,,:f(θ1,θ2,…,θn)=ki=2Σθi.(2)1.2Arm-PSO,。:,。,,。,。,,。PSO,,、,,。,PSO,[8]PSACO,(Arm-PSO).vjt+1=ωtvjt+c1ξ(pjt-θjt)+c2η(pgt-θjt),(3)θjt+1=θjt+vjt+1,(4)ωt=(ωmax-ωmin)×(tmax-t)tmax+ωmin.(5),vjttj,θjt=[θj2,θj2,…,θj6]tj,2~6,pjtj,pgtt,ω,c1c2,ξ、η.,fdis=(p6x-pgx)2+(p6y-pgy)2+(p6z-pgz)2姨.(6)22.1Khatib,.,.[9]GNRON,.,,,.,,,EE(eg,ei,q)=Eatt(eg,q)+l乙Erep(eobsi,q)鄣l=Eatt(eg,q)+ki=1ΣErep(ei,q).(7),q=[x,y,θ]t,eg,eobsii,ei,i=1,…,k,Eatt(eg,q);ki=1ΣErep(ei,q)2022.E(eg,ei,q),F=-荦E=-荦Eatt(eg,q)-ki=1Σ荦Erep(ei,q).(8)E,[9]Eatt(eg,q)=12α1ρ2g(q).(9),α1,ρg(q),ρg(q)=||q-eg||,eg.,,0,。,,ρg(q)≥ρλ,Eatt(eg,q)=αρ2g(q)ρg≥ρλ-α(1/ρg(q)-ε姨+γρg(q)ρg<ρλ姨(10)ρg(q)<ρλ,Eatt(eg,q)=α(1/ρg(q)-ε姨+γ.(11),ρλ=10R,R,γ.,,,,ρg(q)-ε姨,ε,,.[9]Erep(ei,q)=β2(1ρ-ε-1ρ0)2ρmg(q)ifρ(q,qminobs)≤ρ00ifρ(q,qminobs)≤ρ0≤≤≤≤≤≤≤≤≤(12),m,1≤m≤4;ρ(q)q;β;ρ0.:Eatt(eg,q)=-2α(q-qg)ρ≥ρλ-α/(ρ(q)-ε)-γρ≥ρλ≤,(13)F軋rep(ei,q)=βρm(q-qgoal)||q-qminobs||(1ρ-ε-1ρ0)q-qminobsρ+m2(1ρ-ε-1ρ0)2ρm-1(q-qgoal),ifρ(q,qminobs)≤ρ00ifρ(q,qminobs)>ρ0≤≤≤≤≤≤≤≤≤≤≤≤≤≤≤≤≤≤≤≤≤.(14),qminobs.F軋rep=ki=1ΣErep(ei,q).(15)2.2Q-GAα,β,ρ0F=0,.,l.,[10].,,Q-GA.,.Q-GA,Q-GA.[11]Q-GA.:(1),;(2),.(16),fit1,,fit1fit1=alength=astep×l.(16)(17),fit2,,fit2fit2=bρ0.(17),a,b,length,step,l.,getgoal,missgoal.Fit=fit1+fit2,getgoal0,missgoa≤l.(18)3:(1),.(2),.(3).:、、alength=0.35m,.20326awidth=0.35m,aheight=0.2m..w1=0.1m,w2=0.2m,w3=0.05m,w4=0.1m,w5=0.2m,w6=0.05m,h=0.145m.3.1,10m×10m,、x-y,1/2,,2、3。,2.,,。,,3。,,(Experience),,1。1,:(times)(mean)(variance),(LBR);,mean,variance,LBR.3.2,5m×5m×5m,,3.,,,Arm-PSO,.,,,,,θ1=π,θ2=-π/3,θ3=0,θ4=π/2,θ5=π/6,θ6=0.4,1Table1SimulationresultscomparisonofpathplanningMethodTimesLBRMeanVariance1Experience114.7900.61200.0173Q-GA414.4410.84051.501×10-52Experience316.4070.6540.0472Q-GA516.4480.7594.279×10-53Experience510.0100.9580.0626Q-GA510.0001.2191.728×10-64Experience55.6281.9330.1797Q-GA55.6182.1614.33×10-55Experience314.4300.4690.0685Q-GA514.7830.8200.00242Fig.2APFtrapleadstolocalminimumgoalstart3Q-GAAPFFig.3MotionplanningbasedonAPFwithQ-GAstartgoal4Fig.4Globalmotionplanninginthree-dimensionalspace3020100010203001020302042,。3.3、。,,。5,,.5Arm-PSO,200.20,,0.3s1×10-5,1×10-7;5s1×10-5,1×10-14,。2、,210,4000,20.Ts20,N,AC,T。.21×10-4,,.1×10-5,,1-(0.05)2=99.75%.,[6](mountainclimbing,MC),Arm-PSO:1000,10;MC[6],Arm-PSO5.Arm-PSO,,Arm-PSOMC,.N,AC,T.4,,,Q-GA,Arm-PSO..,,,,、.:[1],,.[J].,2003,25(5):465-470.[2]YamamotoY,YunXP.Coordinatinglocomotionandmanipulationofamobilemanipulator[C]//Proceed-ingsofthe31stConferenceonDecisionandcontrol.Arizona,USA:IEEEPress,1992:2643-2648.[3]SerajiH.Motioncontrolofmobilemanipulators[C]//2、Table2Iterationtimes,positioningtimeandaccuracyTsAC/mArm-PSONT/s201×10-2730.032201×10-33560.079201×10-44360.344181×10-59230.7035Fig.5Positioncontrolofmobilemanipulator0.450.400.350.300.250.200.150.10.050010020030040050060070080090010003Arm-PSOMCTable3ComparisonbetweenQ-GAandGAArm-PSOMCNT/sNT/s10.00281240.0781372.3820.0298310.0311011.7630.0031190.0741933.4540.0261370.03511482.6550.000254240.2972564.58AC/m(232).205264BDFEKFFig.4TheratioofwidthbetweenBDFandEKFx10.80.60.40.200102030405060709080100x1:Ratio(BDF/EKF)x1

1 / 7
下载文档,编辑使用

©2015-2020 m.777doc.com 三七文档.

备案号:鲁ICP备2024069028号-1 客服联系 QQ:2149211541

×
保存成功