Motion coordination for a mobile robot in dynamic environments [Elektronische Ressource] / vorgelegt von Boris Kluge

Informations

• Publié par: julius-maximilians-universitat_wurzburg
• Publié le: 01 janvier 2005
• Nombre de lectures: 6
• Langue: English
• Poids de l'ouvrage: 1 Mo

Extrait

MotionCoordination
foraMobileRobot
inDynamicEnvironments
DissertationzurErlangungdes
naturwissenschaftlichenDoktorgrades
derBayerischenJulius Maximilians UniversitätWürzburg
vorgelegtvon
BorisKluge
aus
Creglingen
Würzburg2004Eingereichtam:
beiderFakultätfürMathematikundInformatik
1.Gutachter:
2.Gutachter:
TagdermündlichenPrüfung:ForLuise,Hubert,andUlrike.Abstract
Generating coordinated motion for a mobile robot operating in natural, continuously
changing environments among moving obstacles such as humans is a complex task
which requires the solution of various sub problems. In this thesis, we will cover the
topics of perception and navigation in dynamic environments, as well as reasoning
aboutthemotionoftheobstaclesandoftherobotitself.
Perception is mainly considered for a laser range finder, and an according method for
obstacledetectionandtrackingisproposed. Networkoptimizationalgorithmsareused
for data association in the tracking step, resulting in considerable robustness with re
specttoclutterbysmallobjects.
Navigation in general is accomplished using an adaptation of the velocity obstacle
approach to the given vehicle kinematics, and cooperative motion coordination be
tween the robot and a human guide is achieved using an appropriate selection rule
forcollision freevelocities.
Next,therobotisenabledtocompareitspathtothepathofahumanguideusingoneof
acollectionofpresenteddistancemeasures,whichpermitsthedetectionofexceptional
conditions. Furthermore, a taxonomy for the assessment of situations concerning the
robotispresented,andfollowingasummaryofexistingapproachestomoreintelligent
andcomprehensiveperception,weproposeamethodforobstructiondetection.
Finally,anewapproachtoreflectivenavigationbehaviorsisdescribedwheretherobot
reasons about intelligent moving obstacles in its environment, which allows to adjust
the character of the robot motion from regardful and defensive to more self confident
andaggressivebehaviors.Acknowledgments
I am grateful to all the people who helped me in writing this thesis at various stages.
Special thanks go to Prof. Hartmut Noltemeier, my thesis advisor at the University of
Würzburg for all his support and understanding throughout the work on this thesis.
For being provided with the opportunity to work with inspiring colleagues at an ex
traordinary institute, I am deeply indebted to Prof. Franz Josef Radermacher, head of
the Research Institute for Applied Knowledge Processing (FAW) at the University of
Ulm, and Dr. Thomas Kämpke, head of the autonomous systems department at the
FAW.Especially,numerousdiscussionswithmyformercolleagueChristianSchlegelon
topics ranging (besides robotics) from software engineering to computational geome
tryarethankfullyacknowledged. Finally,Iwanttomentionawonderfultimeworking
withtheentireteamoftheMORPHAproject.
Parts of this work were supported by the German Department for Education and Re
search(BMB+F)undergrantno.01IL902F6aspartoftheprojectMORPHA.Contents
1 Introduction 1
1.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Notations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2 PerceptioninDynamicEnvironments 9
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2 Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.3 FindingObjects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.4 ObjectCorrespondence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.5 Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.6 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3 NavigationinDynamicEnvironments 25
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.2 MobileRobotMotion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.3 CollisionAvoidance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
4 CooperativeMotionCoordination 57
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.2 ProblemDescription . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.3 PracticalMotionCoordination. . . . . . . . . . . . . . . . . . . . . . . . . . 61
4.4 Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63vi CONTENTS
4.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
5 SimilarityofPaths 65
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
5.2 Curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
5.3 DistanceMeasures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
5.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
6 SituationAssessment 115
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
6.2 SituationTaxonomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
6.3 SituationsinCrowdedPublicAreas . . . . . . . . . . . . . . . . . . . . . . 118
6.4 ExistingApproaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
6.5 SituationRecognitioninCrowdedPublicAreas. . . . . . . . . . . . . . . . 122
6.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
7 ReflectiveNavigation 127
7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
7.2 ProbabilisticVelocityObstacles . . . . . . . . . . . . . . . . . . . . . . . . . 128
7.3 RecursiveProbabilisticVelocityObstacles . . . . . . . . . . . . . . . . . . . 135
7.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
7.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
8 Conclusion 161
8.1 PerceptioninDynamicEnvironments . . . . . . . . . . . . . . . . . . . . . 161
8.2 NavigationandMotionCoordination . . . . . . . . . . . . . . . . . . . . . 162
8.3 SituationAssessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
8.4 ReflectiveNavigation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
ListofFigures 167
ListofAlgorithms 169
Bibliography 171