聲音定位技術在許多領域中受到廣泛關注和應用。在工業4.0的智慧工廠中，聲音定位技術能被用於機器故障時的聲源定位，協助工廠實現及時監測和預防維修。在化工4.0的超高壓製程中，以聲音作為高壓氣體微量洩漏的早期偵測機制，能確保工作場所的安全性，提供良好的工作環境。此外還可以應用於機器人定位系統，使機器人能定位聲源來執行相應的任務。麥克風陣列是實現聲音定位的常見方法之一，傳統的麥克風陣列方法主要基於時間差定位，通過測量聲音到達各個麥克風的時間差來計算聲源位置。然而傳統方法的定位精度會受到麥克風數量和陣列物理尺寸的限制。因此減少麥克風的數量和縮小聲音定位系統成為了未來發展的主要目標。
近年來深度學習技術的快速發展為聲音定位帶來了突破。透過使用深度學習，可以從大量的麥克風數據中學習到聲音的響應特徵，實現更準確的定位效果。本論文設計出一種創新單麥克風多通道的聲音定位系統，其主要在於模擬人類耳朵的感知方式。感測範圍能夠達到水平角度0°到360°，以及仰角90°到180°的聲源角度定位。該系統透過不同的通道設計來模擬人類耳廓的功能，進而影響聲源於不同位置時所接收的響應。將不同的響應轉換為圖片，搭配二維卷積神經網路（Two-Dimensional Convolutional Neural Network, 2D CNN）建立模型並與LeNet和AlexNet模型進行比較。結果顯示，該模型在水平角度以及仰角的平均絕對誤差（MAE）分別為8.95°（±2.28°）以及3.18°（±1.24°）。

Acoustic localization technology has received extensive attention and applications in many fields. In the smart factory of Industry 4.0, acoustic localization technology has been used to locate the sound source when the machines fail, in order to assist the factory in real-time monitoring and preventive maintenance. In the Chemical Industry 4.0 ultra-high pressure process, sound is used as an early detection tool for high-pressure gas leaks, which can protect workplace safety. In addition, it can also be applied to the positioning system of the robot, that the robot can accurately perceive and locate the sound source. Microphone arrays are one of the common methods for acoustic localization. The traditional microphone array method is mainly based on time difference localization, and the position of the sound source is calculated by measuring the time difference of sound reaches different microphones in the microphone array. However, the positioning accuracy of traditional methods is limited by the number of microphones and the physical size of the array, and the calculation process is relatively complicated. Therefore, reducing the number of microphones and downsizing the acoustic localization system has become the main goal of future development.
Over the past few years, the rapid development of deep learning technology has brought breakthroughs in sound localization. By using a deep learning, the response characteristics of the sound can be learned from a large amount of microphone data to achieve a more accurate positioning effect. This study designs an innovative single-microphone with multi-channel acoustic localization system, which simulates the way the human ear perceives. The sensing range can reach sound source localization at horizontal angles from 0° to 360°, and elevation angles from 90° to 180°. The system simulates the function of the human auricle through different channel designs, thereby affecting the response received when the sound originates from different locations. Convert different responses into pictures, build a model with a two-dimensional convolutional neural network （2D CNN） and compare with LeNet and AlexNet models. The results show that the Mean Absolute Error（MAE）　of the model in the horizontal angle and elevation angle are 8.95°（±2.28°） and 3.18°（±1.24°）.

[1] X. Zhang, S. Wan, Y. He, X. Wang, L. Dou, Teager energy spectral kurtosis of wavelet packet transform and its application in locating the sound source of fault bearing of belt conveyor, Measurement, 173, 108367, 2021.
[2] D. T. Blumstein, D. J. Mennill, P. Clemins, L. Girod, K. Yao, G. Patricelli, J. L. Deppe, A. H. Krakauer, C. Clark, K. A. Cortopassi, Acoustic monitoring in terrestrial environments using microphone arrays: applications, technological considerations and prospectus, Journal of Applied Ecology, 48(3), 758-767, 2011.
[3] R. C. Luo, C. H. Huang, C. Y. Huang, Search and track power charge docking station based on sound source for autonomous mobile robot applications, 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems, 1347-1352, 2010.
[4] J. D. Harris, R. L. Sergeant, Monaural/binaural minimum audible angles for a moving sound source, Journal of Speech and Hearing Research, 14(3), 618-629, 1971.
[5] Y. LeCun, L. Bottou, Y. Bengio, P. Haffner, Gradient-based learning applied to document recognition, Proceedings of the IEEE, 86(11), 2278-2324, 1998.
[6] A. Krizhevsky, I. Sutskever, G. E. Hinton, Imagenet classification with deep convolutional neural networks, Communications of the ACM, 60(6), 84-90, 2017.
[7] T. A. Rhinehart, L. M. Chronister, T. Devlin, J. Kitzes, Acoustic localization of terrestrial wildlife: Current practices and future opportunities, Ecology and Evolution, 10(13), 6794-6818, 2020.
[8] M.-A. Chung, H.-C. Chou, C.-W. Lin, Sound Localization Based on Acoustic Source Using Multiple Microphone Array in an Indoor Environment, Electronics, 11(6), 890, 2022.
[9] Z. Dong, M. Yu, Research on TDOA based microphone array acoustic localization, 2015 12th IEEE International Conference on Electronic Measurement & Instruments (ICEMI), 1077-1081, 2015.
[10] Q. Yang, Y. Zheng, Deepear: Sound localization with binaural microphones, IEEE Transactions on Mobile Computing, 2022.
[11] X. Sun, H. Jia, Z. Zhang, Y. Yang, Z. Sun, J. Yang, Sound localization and separation in 3D space using a single microphone with a metamaterial enclosure, Advanced Science, 7(3), 1902271, 2020.
[12] A. Saxena, A. Y. Ng, Learning sound location from a single microphone, 2009 IEEE International Conference on Robotics and Automation, 1737-1742, 2009.
[13] 斎藤康毅, Deep Learning：用Python進行深度學習的基礎理論實作, 2017.
[14] Y. Le Cun, L. Bottou, Y. Bengio, Reading checks with multilayer graph transformer networks, 1997 IEEE International Conference on Acoustics, Speech, and Signal Processing, 151-154, 1997.
[15] R. Hazra, S. Majhi, Detecting respiratory diseases from recorded lung sounds by 2d cnn, 2020 5th International Conference on Computing, Communication and Security (ICCCS), 1-6, 2020.
[16] J. Huang, B. Chen, B. Yao, W. He, ECG arrhythmia classification using STFT-based spectrogram and convolutional neural network, IEEE access, 7, 92871-92880, 2019.
[17] J. Middelberg, T. Barber, S. Leong, K. Byrne, E. Leonardi, Computational fluid dynamics analysis of the acoustic performance of various simple expansion chamber mufflers, Proceedings of ACOUSTICS, 123-127, 2004.
[18] A. F. Seybert, D. F. Ross, Experimental determination of acoustic properties using a two‐microphone random‐excitation technique, the Journal of the Acoustical Society of America, 61(5), 1362-1370, 1977.
[19] J. Chung, D. Blaser, Transfer function method of measuring in‐duct acoustic properties. I. Theory, The Journal of the Acoustical Society of America, 68(3), 907-913, 1980.
[20] A. Selamet, P. Radavich, The effect of length on the acoustic attenuation performance of concentric expansion chambers: an analytical, computational and experimental investigation, Journal of Sound and Vibration, 201(4), 407-426, 1997.
[21] E. M. Milad, M. Jolgaf, Acoustic Analysis of a Perforated-pipe Muffler using ANSYS, University Bulletin-Issue(19), 45-56, 2017.
[22] ANSYS, ANSYS WorkBench, URL: https://www.ansys.com/zh-tw/academic/students/ansys-student.
[23] A. Arjunan, C. Wang, K. Yahiaoui, D. Mynors, T. Morgan, V. B. Nguyen, M. English, Development of a 3D finite element acoustic model to predict the sound reduction index of stud based double-leaf walls, Journal of Sound and Vibration, 333(23), 6140-6155, 2014.
[24] S. Pal, Design and acoustic analysis of hybrid muffler, Christ University, 42s, Bangalore, 2015.
[25] 蘇珈正, 微型孔洞聲學超穎材料與吸音口罩應用, 國立中央大學, 2018.
[26] Zhejiang Flashforge 3D Technology, Creator 3 3D Printer, URL: https://www.flashforge.com/product-detail/flashforge-creator-3-pro-3d-printer.
[27] 國家儀器NI, GRAS 40PH-10聲響感測器, URL: https://www.ni.com/zh-tw/support/model.gras-40ph-10.html.
[28] 廖翊涵, 四分之一波長共振腔之寬頻吸音薄板研究, 國立臺灣大學, 2017.
[29] 蔡介峰, 無響室聲場性能驗證與應用之研究 內政部建築研究所, 2017.
[30] 楔石企業有限公司, KEYSTONE聲學纖維方形吸音板, URL: https://www.kphoto.com.tw/index.php?product_id=27196&route=product/product.
[31] UFACTORY, UFACTORY uArm Swift Pro, URL: https://www.cn.ufactory.cc/uarm.
[32] N. Instruments, NI-9234 C系列聲音與振動輸入模組, URL: https://www.ni.com/zh-tw/support/model.ni-9234.html.
[33] 享樂攝影, 自動旋轉拍攝台 URL: https://www.enjoyphoto.com.tw/shop/fotoflex-stptjy300cb/.
[34] Arduino, Arduino Uno, URL: https://store.arduino.cc/products/arduino-uno-rev3.
[35] 陳俊誠, 以 LabVIEW 軟體開發虛擬頻譜分析儀, National Central University, 2009.
[36] M. Dodson, Shannon's sampling theorem, Current science, 63(5), 253-260, 1992.
[37] M. C. Al Fajar, M. Fatmawati, P. Wulandari, D. Astharini, Analysis of dft and fft signal transformation with hamming window in labview, 2020 2nd International Conference on Broadband Communications, Wireless Sensors and Powering (BCWSP), 79-83, 2020.
[38] 曹祐強, 消音器傳輸損失量測方法探討, 國立臺北科技大學, 2016.
[39] D. G. Zill, W. S. Wright, Advanced Engineering Mathematics, Jones & Bartlett Learning, 2014.
[40] 陳振模, 快速傅立葉轉換之複數乘法器設計及其系統整合改良, 國立交通大學, 2005.
[41] J. Zhang, P. T. Lin, Y. Jaluria, Design and optimization of multiple microchannel heat transfer systems, Journal of Thermal Science and Engineering Applications, 6(1), 2014.
[42] W. C. Van Beers, J. P. Kleijnen, Kriging for interpolation in random simulation, Journal of the Operational Research Society, 54, 255-262, 2003.
[43] M. C. E. Manuel, K. S. Hsu, S.-P. Lin, P. T. Lin, User Preference-Oriented Design of Heat Dissipating Elements for Densely Packaged Transistors With Consideration of Design Robustness, Journal of Thermal Science and Engineering Applications, 9(2), 021012, 2017.