Probabilistic Robotics

by Thrun, Sebastian; Burgard, Wolfram; FOX, DIETER

• ISBN13: 9780262201629
• ISBN10: 0262201623
• Format: Hardcover
• Copyright: 2005-08-19
• Publisher: The MIT Press

Summary

Probabilistic robotics is a new and growing area in robotics, concerned with perception and control in the face of uncertainty. Building on the field of mathematical statistics, probabilistic robotics endows robots with a new level of robustness in real-world situations. This book introduces the reader to a wealth of techniques and algorithms in the field. All algorithms are based on a single overarching mathematical foundation. Each chapter provides example implementations in pseudo code, detailed mathematical derivations, discussions from a practitioner's perspective, and extensive lists of exercises and class projects. The book's Web site, http://www.probabilistic-robotics.org, has additional material. The book is relevant for anyone involved in robotic software development and scientific research. It will also be of interest to applied statisticians and engineers dealing with real-world sensor data.

Author Biography

Dieter Fox is Associate Professor and Director of the Robotics and State Estimation Lab in the Department of Computer Science and Engineering at the University of Washington.

Table of Contents

Preface	p. xvii
Acknowledgments	p. xix
Basics	p. 1
Introduction	p. 3
Uncertainty in Robotics	p. 3
Probabilistic Robotics	p. 4
Implications	p. 9
Road Map	p. 10
Teaching Probabilistic Robotics	p. 11
Bibliographical Remarks	p. 11
Recursive State Estimation	p. 13
Introduction	p. 13
Basic Concepts in Probability	p. 14
Robot Environment Interaction	p. 19
Bayes Filters	p. 26
Representation and Computation	p. 34
Summary	p. 35
Bibliographical Remarks	p. 36
Exercises	p. 36
Gaussian Filters	p. 39
Introduction	p. 39
The Kalman Filter	p. 40
The Extended Kalman Filter	p. 54
The Unscented Kalman Filter	p. 65
The Information Filter	p. 71
Summary	p. 79
Bibliographical Remarks	p. 81
Exercises	p. 81
Nonparametric Filters	p. 85
The Histogram Filter	p. 86
Binary Bayes Filters with Static State	p. 94
The Particle Filter	p. 96
Summary	p. 113
Bibliographical Remarks	p. 114
Exercises	p. 115
Robot Motion	p. 117
Introduction	p. 117
Preliminaries	p. 118
Velocity Motion Model	p. 121
Odometry Motion Model	p. 132
Motion and Maps	p. 140
Summary	p. 143
Bibliographical Remarks	p. 145
Exercises	p. 145
Robot Perception	p. 149
Introduction	p. 149
Maps	p. 152
Beam Models of Range Finders	p. 153
Likelihood Fields for Range Finders	p. 169
Correlation-Based Measurement Models	p. 174
Feature-Based Measurement Models	p. 176
Practical Considerations	p. 182
Summary	p. 183
Bibliographical Remarks	p. 184
Exercises	p. 185
Localization	p. 189
Mobile Robot Localization: Markov and Gaussian	p. 191
A Taxonomy of Localization Problems	p. 193
Markov Localization	p. 197
Illustration of Markov Localization	p. 200
EKF Localization	p. 201
Estimating Correspondences	p. 215
Multi-Hypothesis Tracking	p. 218
UKF Localization	p. 220
Practical Considerations	p. 229
Summary	p. 232
Bibliographical Remarks	p. 233
Exercises	p. 234
Mobile Robot Localization: Grid And Monte Carlo	p. 237
Introduction	p. 237
Grid Localization	p. 238
Monte Carlo Localization	p. 250
Localization in Dynamic Environments	p. 267
Practical Considerations	p. 273
Summary	p. 274
Bibliographical Remarks	p. 275
Exercises	p. 276
Mapping	p. 279
Occupancy Grid Mapping	p. 281
Introduction	p. 281
The Occupancy Grid Mapping Algorithm	p. 284
Learning Inverse Measurement Models	p. 294
Maximum A Posteriori Occupancy Mapping	p. 299
Summary	p. 304
Bibliographical Remarks	p. 305
Exercises	p. 307
Simultaneous Localization and Mapping	p. 309
Introduction	p. 309
SLAM with Extended Kalman Filters	p. 312
EKF SLAM with Unknown Correspondences	p. 323
Summary	p. 330
Bibliographical Remarks	p. 332
Exercises	p. 334
The GraphSLAM Algorithm	p. 337
Introduction	p. 337
Intuitive Description	p. 340
The GraphSLAM Algorithm	p. 346
Mathematical Derivation of GraphSLAM	p. 353
Data Association in GraphSLAM	p. 362
Efficiency Consideration	p. 368
Empirical Implementation	p. 370
Alternative Optimization Techniques	p. 376
Summary	p. 379
Bibliographical Remarks	p. 381
Exercises	p. 382
The Sparse Extended Information Filter	p. 385
Introduction	p. 385
Intuitive Description	p. 388
The SEIF SLAM Algorithm	p. 391
Mathematical Derivation of the SEIF	p. 395
Sparsification	p. 398
Amortized Approximate Map Recovery	p. 402
How Sparse Should SEIFs Be?	p. 405
Incremental Data Association	p. 409
Branch-and-Bound Data Association	p. 415
Practical Considerations	p. 420
Multi-Robot SLAM	p. 424
Summary	p. 432
Bibliographical Remarks	p. 434
Exercises	p. 435
The FastSLAM Algorithm	p. 437
The Basic Algorithm	p. 439
Factoring the SLAM Posterior	p. 439
FastSLAM with Known Data Association	p. 444
Improving the Proposal Distribution	p. 451
Unknown Data Association	p. 457
Map Management	p. 459
The FastSLAM Algorithms	p. 460
Efficient Implementation	p. 460
FastSLAM for Feature-Based Maps	p. 468
Grid-based FastSLAM	p. 474
Summary	p. 479
Bibliographical Remarks	p. 481
Exercises	p. 482
Planning and Control	p. 485
Markov Decision Processes	p. 487
Motivation	p. 487
Uncertainty in Action Selection	p. 490
Value Iteration	p. 495
Application to Robot Control	p. 503
Summary	p. 507
Bibliographical Remarks	p. 509
Exercises	p. 510
Partially Observable Markov Decision Processes	p. 513
Motivation	p. 513
An Illustrative Example	p. 515
The Finite World POMDP Algorithm	p. 527
Mathematical Derivation of POMDPs	p. 531
Practical Considerations	p. 536
Summary	p. 541
Bibliographical Remarks	p. 542
Exercises	p. 544
Approximate POMDP Techniques	p. 547
Motivation	p. 547
QMDPs	p. 549
Augmented Markov Decision Processes	p. 550
Monte Carlo POMDPs	p. 559
Summary	p. 565
Bibliographical Remarks	p. 566
Exercises	p. 566
Exploration	p. 569
Introduction	p. 569
Basic Exploration Algorithms	p. 571
Active Localization	p. 575
Exploration for Learning Occupancy Grid Maps	p. 580
Exploration for SLAM	p. 593
Summary	p. 600
Bibliographical Remarks	p. 602
Exercises	p. 604
Bibliography	p. 607
Index	p. 639
Table of Contents provided by Ingram. All Rights Reserved.