title page
Abstract
Contents
List of Abbreviations 13
List of Notations 15
1. Introduction 16
1.1 Motivation 16
1.2 Objective 16
1.3 Major Contribution 17
1.4 Thesis Organization 18
2. Location Estimation Techniques 19
2.1. Distance Measurement Techniques 19
2.2. The Architecture of Indoor Location Systems 20
2.2.1. The Active Mobile Architecture 20
2.2.2. The Passive Mobile Architecture 21
2.3. Comparison of location System 22
2.3.1. In-building RADAR 23
2.3.2. The Active Bat Location System 23
2.3.3. The Active Badge Location System 24
2.3.4. HiBall Head Tracking System 25
2.3.5. Ubisense Location System 25
2.3.6. The Cricket System 26
2.4. Ultrasonic Positioning Techniques 27
2.4.1. Principle 27
2.4.2. Ultrasonic Position Estimation 28
3. A Prototype of Proposed Location System with Accelerometer 32
3.1. Hardware Implementation 32
3.1.1. Detection Method of Ultrasonic Signal 32
3.1.2. Multiple Receiver Sensors 37
3.2. Accelerometer 38
4. Location Estimation Algorithm for High Precision 40
4.1. Statistical Methods 40
4.2. Outlier Rejection Method (ORM) 42
5. Extended Kalman Filters 45
5.1. Principle of an EKF 45
5.1.1. Nonlinear Dynamic Systems 45
5.1.2. Algorithm 47
5.1.3. Kalman Gain 48
5.2. Proposed System Model with EKF 49
5.2.1. Position-Velocity Model 49
5.2.2. Extended Kalman Filter with Sensor Fusion 50
6. Performance Evaluation and Discussion 53
6.1. Experimental Setup 53
6.2. Methodology 56
6.3. Assessment of Location Tracking 57
6.3.1. Location Tracking Experiment in Rectangular Path 57
6.3.2. 1-D Location Tracking Experiment with Accelerometer 63
7. Conclusion and Future Work 68
7.1. Conclusion 68
7.2. Future Work 69
[국문요약] 71
References 73
Acknowledgement 75
Curriculum Vitae 76
3.1. Average error of position estimation at fixed position 36
6.1. The average error and maximum error of raw data 62
6.2. Comparison with P model and PV model 63
2.1. In an active mobile architecture, an active transmitter on… 21
2.2. In a passive mobile architecture, fixed nodes at known positions… 22
2.3. Principle of Tri-lateration 28
2.4. There are two possible listener positions that satisfy the… 29
2.5. The coordinate system used of beacons 29
3.1. Schematics of PLL Detection 33
3.2. Ultrasonic Waveform of PLL Detection 33
3.3. Schematics of Envelop Detection 34
3.4. Ultrasonic Waveform of Envelop Detection 34
3.5. Position estimation with multi-sensor wireless location system in fixed coordinate 36
3.6. (a) Multi-sensor wireless location system and (b) Used three sensor array 37
3.7. ADXL202-Dual Axis Accelerometer kit used in experiment 39
4.1. Flow chart of used algorithms in this thesis 41
4.2. Distance estimation using statistical methods in multi-sensor wireless location system 42
4.3. Flow Chart of Outlier Rejection Method 44
5.1. The flow chart of sensor fusion with EKF 52
6.1. ERI robot with multi-sensor wireless location system 54
6.2. Picture of the expeimental setup 54
6.3. The LEGO train attached multi-sensor listener 55
6.4. Schematic representation of the LEGO train's trajectory 55
6.5. Beacons of the experimental setup 56
6.6. Estimated trajectory with EKF at 0.35m/s 58
6.7. Estimated trajectory with EKF at 0.51m/s 58
6.8. Estimated trajectory with EKF at 0.70m/s 59
6.9. Estimated trajectory with EKF at 0.91mt/s 59
6.10. Error CDF of the different velocities without EKF 60
6.11. P-model error CDF of tile different velocity with EKF 61
6.12. PV model error CDF of the different velocity with EKF 61
6.13. Comparison of error CDF off model and PV model 62
6.14. The measured acceleration of accelerometer at 0.91m/s 64
6.15. Comparison between real path and relative path 64
6.16. Observed path versus estimated position using EKF and sensor fusion at 0.35 m/s 66
6.17. Error CDF of observed path versus estimated position using EKF and sensor fusion at 0.35 m/s 66
6.18. Observed path versus estimated position using EKF and sensor fusion at 0.70 m/s 67
6.19. Error CDF of observed path versus estimated position using EKF and sensor fusion at 0.70 m/s 67