Available:*
Library | Item Barcode | Call Number | Material Type | Item Category 1 | Status |
---|---|---|---|---|---|
Searching... | 30000010280455 | TJ211.415 W365 2011 | Open Access Book | Book | Searching... |
On Order
Summary
Summary
Simultaneous localization and mapping (SLAM) is a process where an autonomous vehicle builds a map of an unknown environment while concurrently generating an estimate for its location. This book is concerned with computationally efficient solutions to the large scale SLAM problems using exactly sparse Extended Information Filters (EIF).
The invaluable book also provides a comprehensive theoretical analysis of the properties of the information matrix in EIF-based algorithms for SLAM. Three exactly sparse information filters for SLAM are described in detail, together with two efficient and exact methods for recovering the state vector and the covariance matrix. Proposed algorithms are extensively evaluated both in simulation and through experiments.
Table of Contents
Preface | p. v |
Acknowledgments | p. vii |
1 Introduction | p. 1 |
1.1 The SLAM Problem and Its Applications | p. 2 |
1.1.1 Description of the SLAM Problem | p. 2 |
1.1.2 Applications of SLAM | p. 3 |
1.2 Summary of SLAM Approaches | p. 4 |
1.2.1 EKF/EIF based SLAM Approaches | p. 5 |
1.2.2 Other SLAM Approaches | p. 8 |
1.3 Key Properties of SLAM | p. 12 |
1.3.1 Observability | p. 12 |
1.3.2 EKF SLAM Convergence | p. 19 |
1.3.3 EKF SLAM Consistency | p. 34 |
1.4 Motivation | p. 41 |
1.5 Book Overview | p. 43 |
2 Sparse Information Filters in SLAM | p. 47 |
2.1 Information Matrix in the Full SLAM Formulation | p. 47 |
2.2 Information Matrix in the Conventional EIF SLAM Formulation | p. 52 |
2.3 Meaning of Zero Off-diagonal Elements in Information Matrix | p. 54 |
2.4 Conditions for Achieving Exact Sparseness | p. 57 |
2.5 Strategies for Achieving Exact Sparseness | p. 59 |
2.5.1 Decoupling Localization and Mapping | p. 59 |
2.5.2 Using Local Submaps | p. 60 |
2.5.3 Combining Decoupling and Submaps | p. 60 |
2.6 Important Practical Issues in EIF SLAM | p. 61 |
2.7 Summary | p. 62 |
3 Decoupling Localization and Mapping | p. 63 |
3.1 The D-SLAM Algorithm | p. 64 |
3.1.1 Extracting Map Information from Observations | p. 64 |
3.1.2 Key Idea of D-SLAM | p. 69 |
3.1.3 Mapping | p. 69 |
3.1.4 Localization | p. 71 |
3.2 Structure of the Information Matrix in D-SLAM | p. 77 |
3.3 Efficient State and Covariance Recover | p. 78 |
3.3.1 Recovery Using the Preconditioned Conjugated Gradient (PCG) Method | p. 80 |
3.3.2 Recovery Using Complete Cholesky Factorization | p. 82 |
3.4 Implementation Issues | p. 84 |
3.4.1 Admissible Measurements | p. 84 |
3.4.2 Data Association | p. 86 |
3.5 Computer Simulations | p. 88 |
3.6 Experimental Evaluation | p. 95 |
3.6.1 Experiment in a Small Environment | p. 95 |
3.6.2 Experiment Using the Victoria Park Dataset | p. 95 |
3.7 Computational Complexity | p. 99 |
3.7.1 Storage | p. 102 |
3.7.2 Localization | p. 102 |
3.7.3 Mapping | |
3.7.4 State and Covariance Recovery | p. 103 |
3.8 Consistency of D-SLAM | p. 107 |
3.9 Bibliographical Remarks | p. 108 |
3.10 Summary | p. 111 |
4 D-SLAM Local Map Joining Filter | p. 113 |
4.1 Structure of D-SLAM Local Map Joining Filter | p. 114 |
4.1.1 State Vectors | p. 115 |
4.1.2 Relative Information Relating Feature Locations | p. 116 |
4.1.3 Combining Local Maps Using Relative Information | p. 116 |
4.2 Obtaining Relative Location Information in Local Maps | |
4.2.1 Generating a Local Map | p. 117 |
4.2.2 Obtaining Relative Location Information in the Local Map | p. 118 |
4.3 Global Map Update | p. 122 |
4.3.1 Measurement Model | p. 122 |
4.3.2 Updating the Global Map | p. 122 |
4.3.3 Sparse Information Matrix | p. 124 |
4.4 Implementation Issues | p. 125 |
4.4.1 Robot Localization | p. 125 |
4.4.2 Data Association | p. 126 |
4.4.3 State and Covariance Recovery | p. 127 |
4.4.4 When to Start a New Local Map | p. 128 |
4.5 Computational Complexity | p. 128 |
4.5.1 Storage | p. 128 |
4.5.2 Local Map Construction | p. 129 |
4.5.3 Global Map Update | p. 129 |
4.5.4 Rescheduling the Computational Effort | p. 130 |
4.6 Computer Simulations | p. 130 |
4.6.1 Simulation in a Small Area | p. 130 |
4.6.2 Simulation in a Large Area | p. 134 |
4.7 Experimental Evaluation | p. 140 |
4.8 Bibliographical Remarks | p. 147 |
4.9 Summary | p. 149 |
5 Sparse Local Submap Joining Filter | p. 151 |
5.1 Structure of Sparse Local Submap Joining Filter | p. 152 |
5.1.1 Input to SLSJF - Local Maps | p. 153 |
5.1.2 Output of SLSJF - One Global Map | p. 154 |
5.2 Fusing Local Maps into the Global Map | p. 155 |
5.2.1 Adding X (K+1)s G into the Global Map | p. 155 |
5.2.2 Initializing the Values of New Features and (K+1)e G in the Global Map | p. 156 |
5.2.3 Updating the Global Map | p. 157 |
5.3 Sparse Information Matrix | p. 158 |
5.4 Implementation Issues | p. 159 |
5.4.1 Data Association | p. 160 |
5.4.2 State and Covariance Recovery | p. 162 |
5.5 Computer Simulations | p. 162 |
5.6 Experimental Evaluation | p. 169 |
5.7 Discussion | p. l69 |
5.7.1 Computational Complexity | p. 169 |
5.7.2 Zero Information Loss | p. 173 |
5 7.3 Tradeoffs in Achieving Exactly Sparse Representation | p. 174 |
5.8 Summary | p. 175 |
Appendix A Proofs of EKF SLAM Convergence and Consistency | p. 177 |
A.l Matrix Inversion Lemma | p. 177 |
A.2 Proofs of EKF SLAM Convergence | p. 178 |
A.3 Proofs of EKF SLAM Consistency | p. 181 |
Appendix B Incremental Method for Cholesky Factorization of SLAM Information Matrix | p. 185 |
B.l Cholesky Factorization | p. 185 |
B.2 Approximate Cholesky Factorization | p. 186 |
Bibliography | p. 189 |