Adaptive navigation and motion planning for autonomous mobile robots [Elektronische Ressource] / vorgelegt von Ashraf Aboshosha

Informations

• Publié par: eberhard_karls_universitat_tubingen
• Publié le: 01 janvier 2004
• Nombre de lectures: 7
• Langue: English
• Poids de l'ouvrage: 6 Mo

Extrait

Acknowledgments
I wish to express my highest appreciation and gratitude to my advisors Prof. Dr.
Andreas Zell and Prof. Dr. Wolfgang Straßer for the encouragement, suggestions
and offering me the facilities necessary for this work.
Appreciation is extended to my current and former colleagues of Computer
Science Dept., Faculty of Informatics, University of Tübingen for their hospitality,
friendship and cooperation. It is a great honour to join this international group.
I learned valuable teamwork within a multi cultural group, how to exchange ideas
with respect among a qualified academic staff and how to explore the world of
science and technologies for new innovations.
I would like to thank S. Wiest, A. Mojaev, H. Tamimi, P. Heinemann, C. Wal-
ter, C. Motoc and K. Beyreuther.
Also, I would like to acknowledge the financial support by the German Aca-
demic Exchange Service (DAAD) of my Ph.D. scholarship at the University of
Tübingen.
I’m indebted to Amal, Mohamed and Osama for their patience in difficult
circumstances and under stress for 4 years to end this work, I’m deeply grateful
to them.
IList of Abbreviations
2D two dimensional
3D three dimensional
AC alternating current
ANFIS adaptive-network-based fuzzy inference systems
ARX auto regressive exogenous
APD avalanche photo-diode
B-spline bases spline
CCA cross correlation analysis
CCD charge couple device
CMOS complementary metal oxide silicon
COG center of gravity
COV covariance
DC direct current
DCT discrete cosine transform
DFT discrete Fourier transform
DST discrete sine transform
DWT Daubechies Wavelets
ED Euclidean distance
FFT fast Fourier transform
FIS fuzzy inference system
FLC fuzzy logic control
GA genetic algorithms
GPS global positioning system
GUI graphical user interface
HSI hue saturation intensity
HSV hue sa value
HWT Haar wavelet transform
IGRF international geomagnetic reference field
IR infrared
LAN local area network
LASER light amplification by simulated emission of radiation
LMS least mean square
IIMIL matrix inversion lemma
MIMO multi input multi output
ML maximum likelihood
MRA multi resolution analysis
MVC model-view-controller
NN neural networks
OO object oriented
RADAR radio ranging
RGB red-green-blue color model
RL reinforcement learning
RLS recursive least squares
PC personal computer
PAC pole assignment control
PCB printed circuit board
PPP probabilistic path planning
SISO single input single output
SLAM simultaneous (concurrent) localization and mapping
SLN straight line navigation
S-line st line
SONAR sound navigation and ranging
TOF time of flight
TS Takagi-Sugeno
Var variance
VLSI very large scale integration
VMP vector mapping paradigm
WMM world magnetic map
IIIList of Symbols
Symbols of Sonar and Laser Integration
r distance
S sonar signal powertrans
f frequency
I(Θ) sonar characteristic radiation function
Θ azimuthal angle of sonar
J Bessel function of the first kind
1
K wave number
λ sonar wave length
m ,m mean value of x, yx y
E() expectation
ρ cross correlation
τ delay
S speed of sounds
l speed of lights
F(ω) spectral analysis
d data
a wavelet averagesi
c coefficientsi
h scaling function coefficientsi
g wavelet fn cotsi
Φ(x) analyzing wavelet or mother wavelet
l location
W(x) scaling function for the mother function
δ delta f
IVSymbols of Geomagnetic Compasses
I inclination
D declination
H horizontal intensity
F total intensity
Z vertical intensity
X north
Y east
SymbolsofMotionPlanningforNon-HolonomicRobots
l path length
Θ deviation angle
nL lagrange polynomiali
nB B-spline polynomiali
Symbols of Image Processing
R red component of RGB model
G green component of RGB model
B blue component of RGB model
H hue component of HSV
S saturation component of HSV model
V value color component of HSV model
ρ cross correlation
Λ guide distance
N number of pixels
VSymbols of Adaptive Navigation
Θ system parameters
Υ aggregated system output
Θ estimated system parameters
υ system output
Φ input/output readings
white noise
C covariance matrix
Ψ Kalman based update matrix
λ system input
−1A(q ) system model poles
−1B(q ) system model zeros
−1T(q ) assigned controller poles
−1G(q ) feedback controller component
−1F(q ) control signal conditioning
−1D(q ) state to output mapping matrix
−1q backward shift operator
H compensatorc
n number of polesa
n number of zerosb
α network learning speed factor
N number of parameters, weights or coefficients
L log likelihood function
V variance
i, j, k counters
S(t) state either υ(t) or λ(t)
o system output
E() expectation or the average
p() probability
VIContents
1 Introduction 1
2 Integration of Distributed Sensors 5
2.1 BenefitsandDrawbacksofSensorIntegration............ 5
2.2 SensorClasification................ 7
2.3 SensorFusion.......... 8
2.4 IntegrationSchemeDesign... 10
2.4.1 Overview .................. 10
2.4.2 RWI-B21RobotPlatform......... 10
2.4.3 MVCbasedSoftwareDevelopment.... 1
3 Sonar Laser Integration 13
3.1 SonarRangingModules ........................ 13
3.1.1 RelevantResearchWork.......... 14
3.1.2 SonarTransducerOperationTheory ... 14
3.1.3 WeaknesesofSonarSensors........ 18
3.1.4 AdvantagesofSonarSensors........... 19
3.2 LaserRangeFinders................ 19
3.2.1 TOFLaserRangingPrinciples....... 19
3.2.2 Strengthsof2DLaserRangeFinders... 21
3.2.3 Problemsof2DLaserRangeFinders............. 21
3.2.4 VectorMappingParadigm(VMP)..... 23
3.2.5 ModelReference .............. 24
3.2.6 RelatedMappingParadigms........ 24
3.3 MatchingofLaserSignatures........... 27
3.3.1 RelevantResearchWork.......... 28
3.3.2 SpatialAnalysis(ED-CCA) ....... 29
3.3.3 SpectralAnalysis(DCT). 31
3.3.4 WaveletAnalysis........................ 34
3.3.5 NotesontheAppliedMatchingTechniques.......... 40
3.4 LaserSonarIntegration..... 41
VII4 Compasses 43
4.1 NaturalMagneticField......................... 43
4.1.1 DefinitionofMagneticElements...... 4
4.1.2 MagneticUnits..... 45
4.1.3 FieldVariations..... 45
4.1.4 CorectionofCompasBearing................ 47
4.2 RobotCompas................... 47
4.2.1 RelatedWork...... 47
4.2.2 FluxgateCompas... 50
4.2.3 MagneticShielding............. 52
4.3 GeomagneticLocalization............. 54
5 Motion Planning for Non-Holonomic Robots 59
5.1 Introduction..................... 59
5.2 RelevantResearchWork.............. 61
5.2.1 Visibility Graph Path Planning . . .... 62
5.2.2 VoronoiDiagram.... 63
5.2.3 PotentialFieldMethod..................... 63
5.2.4 The Probabilistic Path Planner (PPP) ............ 64
5.2.5 BugAlgorithms............... 65
5.3 StraightLineNavigation(SLN)Algorithm.... 6
5.3.1 InitializationPhase............. 66
5.3.2 SegmentationPhase ...................... 66
5.3.3 LinearizationPhase............. 71
5.3.4 MinimizationPhase... 71
5.3.5 RelaxationPhase.... 71
5.4 NotesontheSLNAlgorithm...................... 79
5.5 SLNversusVoronoi................. 81
5.6 SLNversusBug......... 81
6 Adaptive Navigation 85
6.1 RelatedWork.............................. 86
6.2 ModelingofRobotDynamics ........... 87
6.2.1 SignalandSystemModeling........ 88
6.2.2 APrioriandAPosterioriModeling.... 89
6.2.3 ApplicationofAPosterioriModels.............. 89
6.2.4 StochasticKalmanFilter.......... 90
6.2.5 ARXModelingofRobotDynamics.... 92
6.2.6 MatrixInversionLemma(MIL)LearningRule........ 93
6.3 VisualGuideExtraction........................101
6.3.1 SelectionofanAppropriateColorModel...........102
VIII