簡易檢索 / 詳目顯示

研究生: 林逸儒
Lin, I-Ju
論文名稱: 未知水下聲源訊號估測
Estimator Design for Underwater Sound Source with Strong Background Noise Effects
指導教授: 陳永裕
Chen, Yung-Yu
學位類別: 碩士
Master
系所名稱: 工學院 - 系統及船舶機電工程學系
Department of Systems and Naval Mechatronic Engineering
論文出版年: 2017
畢業學年度: 105
語文別: 英文
論文頁數: 90
中文關鍵詞: 卡爾曼濾波器水下聲源自艦噪聲
外文關鍵詞: Kalman filter, Underwater Sound, Self-noise
相關次數: 點閱:62下載:0
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報
  • 本論文提出一個於強烈背景噪音情況下水下聲源的處理方法,此方法是基於多層卡爾曼濾波器的適應性估測設計。本論文設計此多層卡爾曼濾波器的其中一個目的是估測在強烈背景噪音下的原始訊號及系統模型參數,即在自身船艦掛載水下聲源偵測裝置但是自身船艦會產生極為吵雜的自艦噪聲,而此方法的另外一項任務就是用來純化嘈雜的水下聲源。在估測效能測試中,本論文提出的方法在某幾組原始訊雜比低於-8分貝情況下可將水下聲源大幅地提升至可辨識的程度,當訊雜比超過-2分貝代表可辨識,而此方法不論在模擬的結果亦或是實作的結果其提升的訊雜比平均達13分貝以上。本論文提出的估測原始訊號方法有兩個主要的貢獻。首先,本方法採用簡單的水聽器陣列僅由兩個水聽器組成,相對於現有其他常見的方法這也是最為簡易的硬體配置。第二個貢獻則是在受強烈的自艦噪聲影響的情況下,此方法亦可有效的還原接收的聲源訊號。

    An adaptive estimation design which is based on a multi-layer Kalman filter is proposed for the underwater sound recovery in this investigation. One main purpose of this multi-layer design is used to estimate the model parameters and estimate the original signal of background noises, i.e., self-noises come from ships with an underwater sound detector, and the other missions of this proposed method are utilized to purify the noisy underwater sound. From the estimation tests of this proposed method, several noisy underwater sounds which the initial SNRs are all lower than -8 dB can be highly improved to an identifiable level (-2 dB), and an average improved SNR of 13 dB can be obtained no matter simulation results or practical tests. There are two significant contributions delivered from this proposed estimation method. First, a simple hydrophone array with only two hydrophones is used, and this is the simplest arrangement in hardware with respect to the conventional existing designs. Second, this method can effectively recover the underwater sound under a really strong self-noise.

    CONTENTS 中文摘要 I ABSTRACT II 誌謝 III CONTENTS IV LIST OF TABLES VI LIST OF FIGURES VII NOMENCLATURES XIII CHAPTER1 INTRODUCTION 1 1.1 Research Motivation 1 1.2 Literatures Review 2 1.3 Research Methods 3 CHAPTER2 THEORETICAL DERIVATION AND SYSYEM PARAMETERS 5 2.1 Adaptive Estimator Design 5 2.2 Signal-to-Noise Ratio (SNR) 11 CHAPTER3 SIMULATION AND PRACTICAL IMPLEMENTATION 12 3.1 Simulation Results in the Ocean Underwater Environment 12 3.1.1 Simulation Setting 12 3.1.2 Simulation Results 14 3.2 Simulation Results in NCKU Towing Tank 41 3.2.1 Simulation Setting 41 3.2.2 Simulation Results 43 3.3 Practical Experiment in NCKU Towing Tank 56 3.3.1 Experiment Configuration 56 3.3.2 Estimating Source Signals 59 3.3.3 Practical Experiment Results 62 CHAPTER4 CONCLUSIONS 85 REFERENCE 87   LIST OF TABLES Table 1. Comparison between original signal and reducing signal 16 Table 2. Comparison between original signal and reducing signal 23 Table 3. Comparison between original signal and reducing signal 30 Table 4. Reducing performance with different system order in ocean simulation 38 Table 5. Analysis of noise correlation coefficient 54 Table 6. Simulation results in difference combination 55 Table 7. Reducing performance with different system order in Towing tank simulation 55 Table 8. Analysis of noise correlation coefficient 73 Table 9. Analysis of SNR difference 83 Table 10. Reducing performance with different system order in practical 84   LIST OF FIGURES Figure 1. The flowchart of the proposed multi-layer Kalman filter design for the cancellation of self-noises and underwater sound estimation. 11 Figure 2. Profile of sound speed in the underwater environment and arrangements of the underwater sound source, self-noise and hydrophone array 13 Figure 3. Simulation flow chart for constructing the reference signal and the main signal 14 Figure 4. Hydrophone 2 (Main signal) 17 Figure 5. Hydrophone 5 (Reference signal) 17 Figure 6. Parameters histories of the identified first-layer background noises model by the first-layer Kalman filter design 18 Figure 7. Parameters histories of the identified first-layer main sound source model by the second-layer Kalman filter design 18 Figure 8. TLK output in time domain 19 Figure 9. Comparison of TLK output and original underwater sound source 19 Figure 10. TLK output in frequency domain 20 Figure 11. Comparison of TLK output and original underwater sound source in frequency domain 20 Figure 12. SNR evaluation results of applying TLK method and the original underwater sound source along with relative distance 21 Figure 13. Correlation coefficient evaluation results of applying TLK method and the original underwater sound source along with relative distance 21 Figure 14. Improved SNR of applying TLK method and the original underwater sound source along with relative distance along with distance 22 Figure 15. Hydrophone 2 (Main signal) 24 Figure 16. Hydrophone 3 (Refernce signal) 24 Figure 17. Parameters histories of the identified first-layer background noises model by the first-layer Kalman filter design 25 Figure 18. Parameters histories of the identified first-layer main sound source model by the second-layer Kalman filter design 25 Figure 19. TLK output 26 Figure 20. Comparison of TLK output and original underwater sound source 26 Figure 21. TLK output in frequency domain 27 Figure 22. Comparison of TLK output and original underwater sound source in frequency domain 27 Figure 23. SNR evaluation results of applying TLK method and the original underwater sound source along with relative distance 28 Figure 24. Correlation coefficient evaluation results of applying TLK method and the original underwater sound source along with relative distance 28 Figure 25. Improved SNR of applying TLK method and the original underwater sound source along with relative distance along with distance 29 Figure 26. Hydrophone 2 (Main signal) 31 Figure 27. Hydrophone 6 (Reference signal) 31 Figure 28. Parameters histories of the identified first-layer background noises model by the first-layer Kalman filter design 32 Figure 29. Parameters histories of the identified first-layer main sound source model by the second-layer Kalman filter design 32 Figure 30. TLK output 33 Figure 31. Comparison of TLK output and original underwater sound source 33 Figure 32 TLK output in frequency domain 34 Figure 33. TLK output in frequency domain 34 Figure 34. Comparison of TLK output and original underwater sound source in frequency domain 34 Figure 35. SNR evaluation results of applying TLK method and the original underwater sound source along with relative distance 35 Figure 36. Correlation coefficient evaluation results of applying TLK method and the original underwater sound source along with relative distance 35 Figure 37. Improved SNR of applying TLK method and the original underwater sound source along with relative distance along with distance 36 Figure 38. TLK method’s performance with different system order along with distance 39 Figure 39. 3-D plot of SNR difference, noise correlation coefficient and improved SNR 40 Figure 40. Noise correlation coefficient and improved SNR 40 Figure 41. SNR difference and improved SNR 41 Figure 42. Top view of the devices configuration 42 Figure 43. Block diagram of three hydrophones’ receiving signals 43 Figure 44. Hydrophone 2 (Main signal) 44 Figure 45. Hydrophone 1 (Reference signal) 45 Figure 46. Parameters histories of the identified first-layer background noises model by the first-layer Kalman filter design 45 Figure 47. Parameters histories of the identified first-layer main sound source model by the second-layer Kalman filter design 46 Figure 48. TLK output in time domain 46 Figure 49. Noisy underwater sound source and the estimated result 47 Figure 50. Original underwater sound source and the estimated result in time domain 47 Figure 51. TLK output in frequency domain 48 Figure 52. Comparison of TLK output and original underwater sound source in frequency domain 48 Figure 53. Hydrophone 2 (Main signal) 49 Figure 54. Hydrophone 3 (Reference signal) 50 Figure 55. Parameters histories of the identified first-layer background noises model by the first-layer Kalman filter design 50 Figure 56. Parameters histories of the identified first-layer main sound source model by the second-layer Kalman filter design 51 Figure 57. TLK output 51 Figure 58. Noisy underwater sound source and the estimated result 52 Figure 59. Original underwater sound source and the estimated result 52 Figure 60. TLK output in frequency domain 53 Figure 61. Comparison of TLK output and original underwater sound source in frequency domain 53 Figure 62. TLK method’s performance with different system order 56 Figure 63. Concept of the experiment configuration 57 Figure 64. Experiment configuration at coast 57 Figure 65. Concept of the hydrophones’ arrangement 58 Figure 66. Hydrophones’ arrangement on carriage 58 Figure 67. NCKU towing tank background noise 59 Figure 68. Hydrophone 2 60 Figure 69. TLK result compare with original signal 61 Figure 70. TLK output 61 Figure 71. Hydrophone 2 (Main signal) 63 Figure 72. Hydrophone 1 (Reference signal) 64 Figure 73. Parameters histories of the identified first-layer background noises model by the first-layer Kalman filter design 64 Figure 74. Parameters histories of the identified first-layer main sound source model by the second-layer Kalman filter design 65 Figure 75. TLK output 65 Figure 76. Noisy underwater sound source and the estimated result 66 Figure 77. Original underwater sound source and the estimated result 66 Figure 78 TLK output in frequency domain 67 Figure 79. Comparison of TLK output and original underwater sound source in frequency domain 67 Figure 80. Hydrophone 2 (Main signal) 68 Figure 81. Hydrophone 3 (Reference signal) 69 Figure 82. Parameters histories of the identified first-layer background noises model by the first-layer Kalman filter design 69 Figure 83. Parameters histories of the identified first-layer main sound source model by the second-layer Kalman filter design 70 Figure 84. TLK output 70 Figure 85. Noisy underwater sound source and the estimated result 71 Figure 86. Original underwater sound source and the estimated result 71 Figure 87. TLK output in frequency domain 72 Figure 88. Comparison of TLK output and original underwater sound source in frequency domain 72 Figure 89. Hydrophone 3 (Main signal) 74 Figure 90. Hydrophone 1 (Reference signal) 74 Figure 91. Parameters histories of the identified first-layer background noises model by the first-layer Kalman filter design 75 Figure 92. Parameters histories of the identified first-layer main sound source model by the second-layer Kalman filter design 75 Figure 93. TLK output 76 Figure 94. Noisy underwater sound source and the estimated result 76 Figure 95. Original underwater sound source and the estimated result 77 Figure 96. TLK output in frequency domain 77 Figure 97. Comparison of TLK output and original underwater sound source in frequency domain 78 Figure 98. Hydrophone 3 (Main signal) 79 Figure 99. Hydrophone 2 (Reference signal) 79 Figure 100. Parameters histories of the identified first-layer background noises model by the first-layer Kalman filter design 80 Figure 101. Parameters histories of the identified first-layer main sound source model by the second-layer Kalman filter design 80 Figure 102. TLK output 81 Figure 103. Noisy underwater sound source and the estimated result 81 Figure 104. Original underwater sound source and the estimated result 82 Figure 105. TLK output in frequency domain 82 Figure 106. Comparison of TLK output and original underwater sound source in frequency domain 83 Figure 107. TLK method’s performance with different system order 84

    [1] V. O. Knudsen, R. S. Alford, and J. W. Emling, "Underwater ambient noise," J. Marine Res. 7, pp. 410-429, 1948.
    [2] R. Urick and A. Pryce, "A Summary of Underwater Acoustic Data, Part V, Background Noise," Office of Naval Research, Arlington VA,1954.
    [3] G. M. Wenz, "Review of underwater acoustics research: Noise," The Journal of the Acoustical Society of America, vol. 51, no. 3B, pp. 1010-1024, 1972.
    [4] W. M. Carey and R. B. Evans, Ocean ambient noise: measurement and theory. Springer Science & Business Media, 2011.
    [5] R. J. Urick, "Ambient noise in the sea," CATHOLIC UNIV OF AMERICA WASHINGTON DC,1984.
    [6] R. P. Dougherty, "Beamforming in acoustic testing," in Aeroacoustic Measurements: Springer, pp. 62-97, 2002.
    [7] R. P. Dougherty, "Advanced time-domain beamforming techniques," in Proceedings of 10th AIAA/CEAS Aeroacoustics Conference, vol. 10, 2004.
    [8] A. Zhao, B. Zhou, X. Song, and X. Bi, "The technique of diagonal reduction processed with underwater acoustic array signal," J. Harbin Eng. Univ., vol. 35, no. 11, pp. 1327-1331, 2014.
    [9] H. Xia, K. Yang, Y. Ma, S. Zhu, Y. Liu, and Z. Lei, "Study of robust and high-gain beamforming based on diagonal reducing," in OCEANS 2016-Shanghai, pp. 1-5, 2016.
    [10] N. Dal Degan and C. Prati, "Acoustic noise analysis and speech enhancement techniques for mobile radio applications," Signal Processing, vol. 15, no. 1, pp. 43-56, 1988.
    [11] M. M. Goulding and J. S. Bird, "Speech enhancement for mobile telephony," IEEE Transactions on Vehicular Technology, vol. 39, no. 4, pp. 316-326, 1990.
    [12] W. Shen and Y. Ma, "High-resolution bearing estimation for weak signals in nonwhite noise field," in Proceedings of IEEE International Conference. Workshop Marine Acoust., pp. 317-320, 1990.
    [13] D. Zhang, Y. Ma, K. Yang, and Y. Pan, "Weak signal detection method based on sector spatial pre-filtering," in Image and Signal Processing. CISP'08. Congress on, 2008, vol. 5, pp. 552-556, 2008.
    [14] H. Xia, K. Yang, Y. Ma, Y. Wang, and Y. Liu, "Noise Reduction Method for Acoustic Sensor Arrays in Underwater Noise," IEEE Sensors Journal, vol. 16, no. 24, pp. 8972-8981, 2016.
    [15] U. Moreaud, P. Courmontagne, F. Chaillan, J.-R. Mesquida, and S. Ouelha, "A new way for underwater acoustic signal analysis: The morphological filtering," in OCEANS 2015-Genova, pp. 1-9, 2015.
    [16] U. Moreaud, P. Courmontagne, F. Chaillan, and J.-R. Mesquida, "Performance assessment of noise reduction methods applied to underwater acoustic signals," in OCEANS 2016 MTS/IEEE Monterey, pp. 1-15, 2016.
    [17] E. Eweda, "A new performance measure of the normalized LMS adaptive filter," in Signal Processing and Communications, 2007. ICSPC 2007. IEEE International Conference on, pp. 652-655, 2007.
    [18] J. Sanubari, "A new variable step size method for the LMS adaptive filter," in Circuits and Systems, 2004. Proceedings. The 2004 IEEE Asia-Pacific Conference on, vol. 1, pp. 501-504, 2004.
    [19] R. Mehra and R. Kumar, "Improving the SNR of the underwater acoustic signal affected by wind driven ambient noise using RLS algorithm," in Soft Computing Techniques and Implementations (ICSCTI), 2015 International Conference on, pp. 45-50, 2015.
    [20] X. Huang and Y. Wan, "The principle and application of Kalman filtering-MAT LAB Simulation," Beijing: Publishing House of Electronics Industry, pp. 1-5, 2015.
    [21] S. Liu, "An adaptive Kalman filter for dynamic estimation of harmonic signals," in Harmonics and Quality of Power Proceedings, 1998. Proceedings. 8th International Conference On, vol. 2, pp. 636-640, 1998.
    [22] Z. Qifan, Z. Hai, and W. Yanran, "A redundant measurement adaptive Kalman fIlter algorithm," Acta Aeronautiaca et Astronautica Sinica, vol. 36, no. 5, pp. 1596-1605, 2015.
    [23] M. Gabrea, E. Grivel, and M. Najun, "A single microphone Kalman filter-based noise canceller," IEEE Signal Processing Letters, vol. 6, no. 3, pp. 55-57, 1999.
    [24] M. Wu and D. Wang, "A two-stage algorithm for one-microphone reverberant speech enhancement," IEEE Transactions on Audio, Speech, and Language Processing, vol. 14, no. 3, pp. 774-784, 2006.
    [25] E. Hansler and G. Schmidt, Speech and audio processing in adverse environments. Springer Science & Business Media, 2008.
    [26] K. Paliwal and A. Basu, "A speech enhancement method based on Kalman filtering," in Acoustics, Speech, and Signal Processing, IEEE International Conference on ICASSP'87., vol. 12, pp. 177-180, 1987.
    [27] A. Sarafnia and S. Ghorshi, "Noise reduction of speech signal using bayesian state-space Kalman filter," in Communications (APCC), 2013 19th Asia-Pacific Conference on, pp. 545-549, 2013.
    [28] E. Salih, "Speech Enhancement using Kalman filter in communication," Int. J. Math. Comput. Phys. Electr. Comput. Eng, vol. 8, pp. 606-611, 2014.
    [29] M. Mathe, S. P. Nandyala, and T. K. Kumar, "Speech enhancement using Kalman Filter for white, random and color noise," in Devices, Circuits and Systems (ICDCS), 2012 International Conference on, pp. 195-198, 2012.
    [30] C. Pandey and S. Nemade, "Enhancement of the speech quality by the implementation of Second Order Fast Adaptive Kalman Filter algorithm," in India Conference (INDICON) Annual IEEE, 2014, pp. 1-6, 2014.
    [31] A. P. Nair, S. Krishnan, and Z. Saquib, "MFCC based noise reduction in ASR using Kalman filtering," in Conference on Advances in Signal Processing (CASP), pp. 474-478, 2016.
    [32] M. Fujimoto and Y. Ariki, "Noisy speech recognition using noise reduction method based on Kalman filter," in Acoustics, Speech, and Signal Processing, 2000. ICASSP'00. Proceedings. 2000 IEEE International Conference on, vol. 3, pp. 1727-1730, 2000.
    [33] M. Jeub, C. Herglotz, C. Nelke, C. Beaugeant, and P. Vary, "Noise reduction for dual-microphone mobile phones exploiting power level differences," in Acoustics, Speech and Signal Processing (ICASSP), 2012 IEEE International Conference on, pp. 1693-1696, 2012.
    [34] C. M. Nelke, C. Beaugeant, and P. Vary, "Dual microphone noise PSD estimation for mobile phones in hands-free position exploiting the coherence and speech presence probability," in Acoustics, Speech and Signal Processing (ICASSP), 2013 IEEE International Conference on, pp. 7279-7283, 2013.
    [35] W. Nabi, N. Aloui, and A. Cherif, "Speech enhancement in dual-microphone mobile phones using Kalman filter," Applied Acoustics, vol. 109, pp. 1-4, 2016.
    [36] W. Nabi, A. Noureddine, and A. Cherif, "A dual-microphone noise reduction algorithm for mobile communications," in Sciences of Electronics, Technologies of Information and Telecommunications (SETIT), 2016 7th International Conference on, pp. 258-262, 2016.

    無法下載圖示 校內:2022-07-28公開
    校外:不公開
    電子論文尚未授權公開,紙本請查館藏目錄
    QR CODE