Multi-Sensor Data Fusion with MATLAB

【实例简介】
多传感器信息融合的书籍，一本关于多传感器信息融合的经典书籍。
MATLAB and Simulink" are trademarks of The Math Works, Inc and are used with permission. The MathWorks does not warrant the accuracy of the text of exercises in this book. This book's use or dis cussion of MATLAB and Simulink software or related products does not constitute endorsement or sponsorship by The MathWorks of a particular pedagogical approach or particular use of the matlab and Simulink soltware CRC Press Taylor Francis grol 6000 Broken Sound Parkway nw, suite 300 Boca raton Fl 33487-2742 9 2010 by Taylor and Francis Group, LLC CRC Press is an imprint of Taylor Francis Group, an Informa business No clain Lo original U.S. Government works Printed in the United States of America on acid-free paper 10987654321 International Standard Book Number 978-1-4398-0003-4(Hardback) This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been made to publish reliable data and information, but the author and publisher cannot assume copyright holders if permission to publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let us know so we may rectify in any future reprint Except as permitted under U.S. Copyright L aw, no part of this book may be reprinted reproduced, ans- mitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers Torpermissiontophotocopyorusematerialelectronicallyfromthisworkpleaseaccesswww.copyright (http:/www.copyright.com/)orcontacttheCopyrightClearanceCenterInc.(ccc),222Rosewood Drive, Danvers, MA01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe Library of Congress Cataloging-in-Publication Data Raol, R. tendra r. ,1947. Multi-sensor data fusion with MATLAB Jitendra R Raol P·cm CRC title Includes bibliographical references and index ISBN 978-1-4398-0003-4(hardcover: alk. paper 1. Multisensor data fusion-Data processing 2. MATLAB. 3. Detectors. I. Title TA331R362010 681.2-dc22 2009041607 Visit the Taylor francis Web site http://www.taylorandfrancis.com and the crc Press Web site at http://www.crcpress.com o 2010 by Taylor and Francis Group LLC The book is dedicated in loving memory to Professor P N. Thakre (M.S. University of Baroda, Vadodara) Professor Vimal K Dubey (Nanyang Technological university, Singapore), d Professor Vinod modi (University of British Columbia, Canada) o 2010 by Taylor and Francis Group LLC Contents reface……………………………………………… IX Acknowledgments XXI Auth XLll Contributors Introduction..… XXVII Part I: Theory of Data Fusion and Kinematic- Level Fusion ( R Raol, G. Girija, and N. Shanthakuar 1. Introduction… 2. Concepts and Theory of Data Fusion…,,… 11 2.1 Models of the Data fusion process and architectures 2.1.1 Data Fusion models 2.1.1.1 Joint Directors of Laboratories Model 业垂 2.1.1.2 Modified Waterfall Fusion Model..........17 2.1.1.3 Intelligence Cycle-Based mode 18 2.1.1.4 Boyd Model 19 2.1.1.5 Omnibus model ,20 2.1.2 Fusion architectures. ...................... 21 2.1.2.1 Centralized Fusion 21 2.1.2.2 Distributed Fusion ....................................................21 2.1.2.3 Hybrid Fusion 2.2 Unified Estimation Fusion models and Other methods 23 2.2.1 Definition of the Estimation Fusion Process......... 24 22.2 Unified Fusion Models Methodology…………… 25 2.2.2.1 Special Cases of the Unified Fusion Models..25 2.2.2.2 Correlation in the Unified fusion models,,.... 26 2.2. 3 Unified Optimal Fusion Rules.........27 2.2.3.1 Best Linear Unbiased Estimation Fusion rules with Complete prior Knowledge .27 2.2.3.2 Best Linear Unbiased estimation Fusion rules without prior Knowledge 2.2.3.3 Best Linear unbiased estimation Fusion rules with Incomplete prior Knowledge 28 2.2.3.4 Optimal-Weighted Least Squares Fusion Rule..28 2.2.3.5 Optimal Generalized Weighted Least Squares Fusion rule. ..........................................................29 o 2010 by Taylor and Francis Group LLC Contents 2.2.4 Kalman Filter Technique as a Data Fuser…………………29 2.2.4.1 Data Update algorithm 2.2.4.2 State-Propagation Algorithm...... 31 2.2.5 Inference Methods 32 2.2.6 Perception, Sensing, and Fusion.......32 2.3 Bayesian and Dempster-Shafer Fusion Methods 2.3.1 Bayesian Method. 34 2.3.1.1 Bayesian Method for Fusion of Data from Two Sensors……...36 2.3.2 Dempster-Shafer method 2.3.3 Comparison of the bayesian Inference Method and the dempster-Shafer Method 2.4 Entropy-Based Sensor Data Fusion Approach....... 41 2.4.1 Definition of Information 2.4.2 Mutual Information. .......................................43 2.4.3 Entropy in the Context of an Image 2.4.4 Image-Noise Index 44 2.5 Sensor Modeling, Sensor Management, and Information Pooling 2.5.1 Sensor Types and Classification…… 45 2.5.1. 1 Sensor Technology .46 2.5.1.2 Other Sensors and their Important Features and us sages∴… 48 2.5.1.3 Features of sensors 51 2.5.1. 4 Sensor characteristics 52 2.5.2 Sensor management….,,,..………153 2.5.21 Sensor Modeling…………….55 2.5.2.2 Bayesian Network Model......58 2.5.2.3 Situation Assessment Process.............58 2.5.3 Information-Pooling methods 2.5.3.1 Linear Opinion Pool 2.5.3.2 Independent Opinion Pool 2.5.3.3 Independent likelihood pool……….61 3. Strategies and Algorithms for Target Tracking and Data Fusion∴....63 3.1 State-Vector and measurement-Level fusion 3.1.1 State-Vector Fusion 3.1.2 Measurement Data- Level fusion.…….…71 3. 1. 3 Results with Simulated and Real Data Trajectories....71 3.1.4 Results for Data from a Remote Sensing Agency with Measurcment Data-Level fusion 3.2 Factorization Kalman Filters for Sensor Data Characterization and Fusion 3.2.1 Sensor Bias errors o 2010 by Taylor and Francis Group LLC Contents 3.2.2 Error State-Space Kalman Filter 75 3.2.3 Measurement and Process Noise Covariance Estimation… 3.2.4 Time Stamp and Time delay errors 3.2.5 Multisensor Data Fusion Scheme 3.2.5.1 UD Filters for Trajectory Estimation ..................80 3.2.5.2 Measurement fusion 3.2.5.3 State-Vector Fusion 32.54 Fusion Philosophy…………………… 82 3.3 Square-Root Information Filtering and Fusion in Decentralized architecture..........................86 3.3.1 Information Filter 87 3.3.1.1 Information Filter Concept……87 3.3.1.2 Square Root Information Filter Algorithm 3.3.2 Square Root Information Filter Sensor Data Fusion Al igorithm… 88 3.3.3 Decentralized square Root Information Filter 89 3.3.4 Numerical Simulation Results 3.4 Nearest Neighbor and Probabilistic Data Association Filter...91 Algorithms 34.1 Nearest Neighborhood Kalman Filter……………….94 3.4.2 Probabilistic Data Association Filter 3.4.3 Tracking and Data Association Program for Multisensor, multitarget Sensors 97 3.4.3.1 Sensor Attributes....................99 3.4.3.2 Data Set Conversion....................99 3.4.3.3 Gating in Multisensor, Multitarget 100 3.4.3.4 Mcasurcment-to- Track association….……………100 3.4.3.5 Initiation of Track and Extrapolation of Track..101 3.4. 3.6 Extrapolation of Tracks into Next Sensor Field of view…101 34.3.7 Extrapolation of Tracks into Next Scan………102 343.8 Track management process…… 102 3.4.4 Numerical simulation.......................103 3.5 Interacting Multiple Model Algorithm for Maneuvering Target Tracking 3.5.1 Interacting Multiple Model Kalman Filter Algorithm ...106 3.5.1.1 Interaction and mixing .........................108 3.5.1.2 Kalman Filtering………………108 3.5.1.3 Mode Probability Update .109 3.5.1.4 State Estimate and Covariance Combiner. .............10 3.5.2 Target Motion Models………… ∴10 3.5.2.1 Constant Velocity model .110 3.5.2.2 Constant Acceleration model.............110 o 2010 by Taylor and Francis Group LLC Contents 3.5.3 Interacting Multiple Model Kalman Filter Implementation………… 111 3.5.3. 1 Validation with Simulated Data. ............................112 3.6 Joint Probabilistic Data Association Filter 116 3.6.1 General Version of a Joint Probabilistic Data Association filter 117 3.6.2 Particle Filter Sample-Based Joint Probabilistic Data Association filter…… 119 3.7 Out-of-Sequence Measurement Processing for Tracking....120 3.7.1 Bayesian Approach to the Out-of-Sequence Measurement problem. .................................................120 3.7.2 Out-of-Sequence Measurement with Single Delay and No Clutter…...............121 372.1 Y Algorithn……… 121 3.7.2.2 Augmented State Kalman Filters 122 3.8 Data Sharing and Gain Fusion Algorithm for Fusion…………124 3.8.1 Kalman Filter-Based Fusion Algorithm. .124 38.2 Gain Fusion-Based Algorithm………………………125 3.8.3 Performance Evaluation.…….. 126 3.9 Global Fusion and H-Infinity Filter-Based Data Fusion 127 391 Sensor Data Fusion using H-Infinity Filters……………127 3.9.2 H-Infinity a posteriori Filter-Based Fusion Algorith …130 3.9.3 H-Infinity Global Fusion Algorithm 131 3.9.4 Numerical Simulation results ……132 3.10 Derivative-Free Kalman Filters for Fusion ......................................134 3.10.1 Derivative-Free Kalman Filters 136 3.10.2 Numerical Simulation...................137 3.10.2. 1 Initialization of the Data Fusion-Derivative Free Kalman Filter algorithm 140 3.10.2.2 Computation of the Sigma points………,140 3.10.2.3 State and Covariance propagation.…………141 3.10.2. 4 State and Covariance update 141 3.11 Missile seeker estimator 143 3.11.1 Interacting Multiple model-Augmented Extended Kalman Filter algorithm.....….…....143 3.11.1.1 State Model 144 3.11.1.2 Measurement Model 14 3.11.2 Interceptor-Evader Engagement Simulation........ 146 3.11.2. 1 Evader Data Simulation .147 3.11.3 Performance Evaluation of Interacting Multiple model-Augmented Extended Kalman Filter 147 3.12Ⅲ ustrative Examples……………………………151 o 2010 by Taylor and Francis Group LLC Contents 4. Performance Evaluation of Data Fusion Systems Software, and Tracking……………………157 4.1 Real-Time Flight Safety Expert System Strategy…………….160 4.1.1 Autodecision Criteria 161 4. 1. 2 Objective of a Flight Test Range......161 4.1.3 Scenario of the Test Range.………… 161 4.1.3.1 Tracking Instruments .162 4.1.3.2 Data Acquisition 163 4.1.3.3 Decision Display System………163 4.1. 4 Multisensor Data Fusion System .............................163 4.1.4.1 Sensor Fusion for range Safety computer...........164 4.1.4.2 Algorithms for Fusion 164 4.1.4.3 Decision Fusion 165 4.2 Multisensor Single-Target Tracking ·⊥6 4.2.1 Hierarchical Multisensor Data Fusion architecture and 166 4.2.2 Philosophy of Sensor Fusion 168 4.2.3 Data Fusion Software Structure. ..........................................169 4.2.3.1 Fusion module 1 169 4.2.3.2 Fusion modules 2 and 3 .169 42.4 Validation… 170 4. 3 Tracking of a Maneuvering Target-Multiple-Target Tracking Using Interacting Multiple Model Probability Data Association Filter and Fusion 431 Interacting Multiple Model Algorithm………….…….171 4.3.1.1 Automatic track formation 4.3.1.2 Gating and Data Association 172 4.3.1.3 Interaction and Mixing in Interactive Multiplo Model Probabilistic Data Association Filter.... 174 4314 Mode-Conditioned Filtering…… 174 43.1.5 Probability computations…… 175 4.3. 1.6 Combined state and Covariance Prediction and estimation 4.3.2 Simulation validation 177 4.3.2.1 Constant Velocity model 17 4.3.2.2 Constant Acceleration model.............178 4.3.2.3 Performance Evaluation and Discussions. .............179 4.4 Evaluation of Converted Measurement and modified Extended Kalman Filters.................183 4.4.1 Error model Converted measurement Kalman filter and error model modified Extended Kalman filter Algorith 184 4.4.1.1 Error Model Converted measurement Kalman Filter Algorith 185 o 2010 by Taylor and Francis Group LLC Contents 4.4.1.2 Error model modified extended Kalman filter Algorithm................................186 4.4.2 Discussion of results 189 4.4.2.1 Sensitivity Study on Error Model Modified Extended Kalman Filter..............191 4.4.2.2 Comparison of Debiased Converted Measurements kalman filter error model Converted Measurcment Kalman Filter, and Error Model modified Extended Kalman Filter 4.5 Estimation of Attitude Using Low-Cost Inertial Platforms anl.191 Algorithms Kalman filter Fusion 4.5.1 Hardware System 195 45.2 Sensor Modeling…………… 195 4.5.2.1 Misalignment Error Model 196 4.5.2.2 Temperature Drift Model 196 4.5.2.3 CG Offset Model .196 4.5.3 MATLAB/ Simulink Implementation…………,19 4.5.3.1 State Model 197 4.5.3.2 Measurement model 4.5.4 Microcontroller Implementation ..........................200 Epilogue............203 E xercises∴ 203 References 206 Part II: Fuzzy Logic and Decision Fusion ( R. Raol and S K. Kashyap) 5. Introduction 215 6. Theory of Fuzzy Logic…..……….….…217 61 Interpretation and Unification of Fuzzy Logic Operations………218 6.1.1 Fuzzy sets and membership functions 218 6.1.2 Types of Fuzzy Membership Functions 220 6.1.2.1 Sigmoid-Shaped Function...... 220 6.1.2.2 Gaussian-Shaped Function 220 6. 1.2.3 Triangle-Shaped Function 222 6.1.24 Trapezoid-Shaped Function……2 6.1.2.5 S-Shaped function 222 6..2.6∏- Shaped Function…….…2241 6.1.27 Z-Shaped Function…….24 6.1.3 Fuzzy Set Operations 225 6.1.3.1 Fuzzy logic operators 226 o 2010 by Taylor and Francis Group LLC 【实例截图】
【核心代码】