非完整約束條件下移動機器人路徑規(guī)劃.pdf

1、大連理工大學碩士學位論文非完整約束條件下移動機器人路徑規(guī)劃姓名：鄭效光申請學位級別：碩士專業(yè)：模式識別與智能系統(tǒng)指導教師：莊嚴20071201大連理工大學碩士學位論文MobileRobotPathPlanningunderNonholonomicConstraintsAbstractWithmoreandmoreapplicationsinvariousarOfl$，thenonholonomicmobilerobothasb(x20m

2、eafocusinthefieldofroboticsThispaperemphasizes011nonholonomicmobilerobot’Spathplanningfiomtheaspectsoftheoryandapplicationandaseriesofpathplanningalgorithmsarealsoglvcntoadapttodifferentenvironmentwhichcouldmakethemobile

3、robotaccomplishthepathplarmingtaskwiththeabilityofautonomousenvironmentad印tm呂AglobalpathplannerbasedongridpointrepresentationisproposedtoworkintheknownenvironmentNotonlyisanimproveddistancetransformalgorithmusedtogenerat

4、ethedesiredholonomicpathbutalsoisthecubicsplineinterpolationmethodusedtosmoothitThentheA4methodisexploitedtosearchthevalidpathinthenonholonomicroadmapwhichisconsuuetedbystatelatticescomposedofgridpointsThemulti—layerstat

5、elatticesalgorithmispresentedforthespecialenvironmentwitllvariousnanowp姍gesSOthatthenumberofstatelatticenearbytheobstaclesinereasesdramaticallyexceptthefleespaceApathplannerbaseduponoptimizationispropos。d，whichtakesfulla

6、ccountofdifferentimpactsincludingthedistanceandanglebetweentherobotandthetargetthemobilerobot’SvelocityandobstaclesThispathplanningproblemcouldbetransformedtoadiscreteoptimizationproblembyusingofthenavigationevaluationfu

7、nctionInordertodeterminebotlltherangeofthesearchingspeedandoptimalspeedofobstacleavoidancethedynamicwindowmethodandthegeneticalgorithmareusedrespectivelyAhybridpathplannercombiningglobaloptimizationandquickobstacleavoida

8、nceisproposedinpartiallyknownenvironmentTheglobalpathplannerbasedonimproveddistancetransformalgorithmcarlgenerateaseriesofsubgoalpointstothetargetwhicharefarfromtheobstaclesAccordingtotheinformationofthelaserscanningthes

9、teeringcontrolmethodisusedasthelocalpathplannerwhichcouldavoidtheunknownobstaclesprompflySimulationsandexperimentresultsincomplexindoorenvironmentsshowthemethod’seffectivenessandpracticabilityInthelastpartofthispaper，the

10、POMDPnavigationisstudiedanddiscussed，especiallyforthestateestimationandactionselectionin恤edecisionprocessThisworkisverycriticalfortheSut∞essofthePOMDPnavigationinthefutureworkKeyWords：NonholonomicConstraints；Thrccdimaxsi