臺灣博碩士論文加值系統

本論文提出一套在未知聲源個數的情況下能同步定位移動式機器人平台聲源的方法，並解釋了使用聲音作為定位地標的原因。文中介紹了幾種角度估測(DOA)的方法，並結合其中一些方法的特性來達到多聲源角度估測的目的，這些角度估測將使用在理論基礎為貝氏濾波器(Bayes Filter)的純角度資訊同步定位演算法(Bearings-only SLAM)。因聲源非同步發聲，且聲源沒有角度以外的資訊，故有未知資料群落(Unknown Data Association)的問題需解決，演算法中的粒子濾波器(particle filter)將可處理這部分的問題。本研究為了將演算法即時實現在實驗環境中而做了不影響理論結果的修改，並且展示了實驗成果來證實演算法的效能。

This work proposes a method that is able to simultaneously localize a mobile robot and unknown number of multiple sound sources in the environment. The reason of using sound sources as the landmarks in SLAM algorithm is presented. Several DOA estimation methods are described and a combinational one is used for real time application. After knowing the DOA information, a bearings-only SLAM (simultaneous localization and mapping) algorithm is introduced in detail, which contains the theoretical structure of Bayes filter. The estimated DOAs are known as the bearings information in the algorithm. As source signals are not persistent and there is no identification of the signal content, data association is unknown which is solved using particle filter. Modifications of the algorithm are made for real time application. Experimental results are presented to verify the effectiveness of the proposed approaches.

摘要 i
ABSTRACT ii
誌謝 iii
CONTENT iv
LIST OF TABLES vi
LIST OF FIGURES vii
Chapter 1. INTRODUCTION 1
1.1 Motivation and Objective 1
1.2 Literature Review 2
1.3 Thesis Subject and Contribution 3
1.4 Outlines of Thesis 4
Chapter 2. DETECTION OF MULTIPLE SOUND SOURCES 5
2.1 Introduction 5
2.2 Mathematical Structure of TDE and Eigenspace Method 5
2.2.1 TDE 5
2.2.2 ES-GCC, a combination of TDE and Eigenspace Method 7
2.3 Performance of Different Eigenspace Methods in Real Time Application 11
Chapter 3. BEARINGS-ONLY SLAM ALGORITHM 12
3.1 Between DOA estimation and Bearings-Only SLAM 12
3.2 Introduction of Probabilistic Robotics 12
3.3 Bayes Filter 13
3.3.1 State Estimation using Probabilistic Generative Laws 13
3.3.2 Parametric Filter – Kalman Filter and Extended Kalman Filter 15
3.3.3 Nonparametric Filter - Particle Filter 19
3.4 SLAM 21
3.4.1 Problem Definition 21
3.4.2 FastSLAM 23
3.4.3 Unknown Data Association 25
3.5 Simulation Result of Fast SLAM 25
Chapter 4. Experimental Result and Discussion 29
4.1 Introduction of the Robot Platform 29
4.2 Performance of offline calculated algorithm 31
4.3 Performance of online algorithm 34
Chapter 5. Conclusion and Future Study 38
REFERENCE 39

[1] D. Johnson and D Dudgeon, “Array Signal Processing: Concepts and Techniques,” Prentice Hall, Englewood Cliffs, New Jersey, 1993.
[2] Chia-Hsing Yang, “A Real-time Speech Purification and Voice Activity Detection System Using Microphone Array” National Chiao Tung University, Institute of Electrical and Control Engineering, Master Thesis, 2008
[3]. C. H. Knapp and G. C. Carter, “The generalized correlation method for estimation of time delay,” IEEE Transaction on Acoustic Speech, Signal Processing, ASSP-24(4):320-327, Aug. 1976.
[4]. R. O. Schmidt, “Multiple Emitter Location and Signal Parameter Estimation,” IEEE Transaction Antennas and Propagation, vol. AP-34, no. 3, pp.276-280, March 1986
[5]. G.. C. Carter, A. H. Nuttall, and P. G. Cable, “The smoothed coherence transform,” Proc. IEEE (Lett.), vol. 61, pp. 1497-1498, Oct. 1973.
[6] S. Thrun, W. Burgard, D. Fox, Probabilistic Robotics, The MIT Press, Cambridge, 2005.
[7] R. E. Kalman, “A New Approach to Linear Filtering and Prediction Problems,” Transactions of the ASME-Journal of Basic Engineering, 82 (Series D): 35-45, 1960
[8] Cheng-Kang Wang, “Multiple Sound Source Direction Estimation and Sound Source Number Estimation,” National Chiao Tung University, Institute of Electrical and Control Engineering, Master Thesis, 2008
[9] Source Website: http://www.activrobots.com/ROBOTS/specs.html