請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/68849
標題: | 麥克風陣列結合波束成型理論進行聲源定位及應用聲學攝影機於加工系統之振噪分析 Using Microphone Arrays with Beamforming in Sound Source Localization and Application of Sound Camera for Noise and Vibration Analysis of Machining Systems |
作者: | Tzu-Yin Su 蘇紫茵 |
指導教授: | 馬劍清(Chien-Ching Ma) |
關鍵字: | 聲源定位,麥克風陣列,波束成型,聲源可視化,振噪分析, sound source localization,microphone array,beamforming,acoustic source visualizing,noise and vibration analysis, |
出版年 : | 2020 |
學位: | 碩士 |
摘要: | 就像人有兩個耳朵一樣,多個感測器組成陣列後可以進行聲源定位。此技術應用廣泛,從早期用於軍事科技,到現代智慧聲控裝置或輔助工業噪音檢測,皆能見到其蹤影,亦是一項發展已久的技術,研究者也努力從陣列排列方式或演算法優化追求提升定位精準度及運算速度。 本研究首先將聲音感測器以直線的方式排列建立一維等間距線性陣列系統,並使用波束成型演算法進行聲源定位,進而建構完整的自動聲源追蹤系統;接著為了追蹤空間中的聲源位置,以一維陣列系統為基礎建立二維的聲源定位陣列,利用不同的排列形式設計各種陣列幾何的配置,並探討對應的聲源辨識能力與解析度,進而找出最佳的陣列排序方式;接著將二維陣列系統與工業相機結合,建立出可視化的聲場量測系統,並設計靜止、二維移動和三維空間移動的聲源實驗,測試此系統的靈敏度與定位能力;最後將本系統與三維數位影像相關法(digital image correlation,DIC)結合可精準定位聲源的三維空間位置,並將分析結果與市售的聲學攝影機相互比對以驗證本系統之可行性。另一部份則是將聲源定位應用於振噪分析的實例研究,與傳統量測工具相互搭配,針對內藏式主軸、馬達加工系統與機械手臂三種不同加工系統進行量測,將運轉訊號分離出環境背景噪音、電子元件噪音與運轉機械噪音並分項討論,也從異常訊號特徵推論結構有零件鬆動、基座不穩等問題,並成功使用聲學攝影機找到發出異音的缺陷軸承;最後則將聲學攝影機拍攝結果相互比較,討論聲源定位技術於實例應用的可行性。 Similar to human’s two ears, the sound can be “heard” and localized by the sensor array. This technology is widely used in many fields from military technology at early stages to smart voice control devices or assisting industrial noise detection. This technique has been developed for a long time. Researchers usually focus on the topics for the location of those sensors that were placed or developing the algorithm in order to improve locating accuracy, precision, and speed. This study starts with 1-D array system with uniform arrangement, using beamforming algorithm in sound source localization and constructs an automatic source tracking system. For tracking the sound source in space, it carries on 2-D array system and discussed different geometric forms of the array with their locating ability and resolution, therefore finding out the best arrangement. Next combines this 2-D system with industrial camera to practicing an acoustic source visualizing system and then tests its sensitivity and locating ability by tracking a steady sound source, moving in a plane and in space. Lastly, combines this 2-D sound source tracking system with 3-D digital image correlation can locate the sound source in a 3-D space and compare the experimental results with those from a commercially available product, sound camera, to evaluate the system. The other part of this thesis is noise and vibration analysis, which cooperates this sound source localization technology with traditional measurement tools such as microphone and accelerometer. In these 3 cases, which include built-in high-speed spindle, machining system, and robotic arm, this study separates the operation signal for background noise, electronic components noise and mechanical noise from the machines. Finally, the capability of the practical application of sound source localization technology is discussed by the measuring results of the sound camera in these cases. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/68849 |
DOI: | 10.6342/NTU202003666 |
全文授權: | 有償授權 |
顯示於系所單位: | 機械工程學系 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
U0001-1708202003533600.pdf 目前未授權公開取用 | 22.68 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。