可視化噪聲源識別技術

李加慶 陳進 張桂才 雷宣揚 薛偉飛
上海交通大學振動、沖擊、噪聲國家重點實驗室,上海200240

摘　要:簡要回顧了噪聲源識別的傳統方法和基于信號處理的方法，闡述了可視化聲源識別技術的基本原理，然后利用各種重建算法在三維空間中進行聲場重構并將結果用圖形表示出來。按噪聲信號采集方式、重建算法和結果表示方式對該技術進行了分類。對常用的三類重建算法Fourier方法、Helmholtz—Kirchhoff積分方程法和等效源法進行了詳細的比較，重點分析了等效源法，指出它目前存在的研究困難。給出了一種應用可視化噪聲源識別技術的系統解決方案。[著者文摘]

關鍵詞:噪聲源識別 可視化 Fourier方法 Helmholtz—Kirchhoff積分方程法 等效源法

Visualization for noise sources identificationLI Jia-qing, CHEN Jin, ZHANG Gui-cai, LEI Xuan-yang, XUE Wei-fei State Key Lab of Vibration, Shock and Noise, Shanghai Jiaotong University, Shanghai 200240, ChinaAbstract:The principle of visualization techniques on noise sources identification （NSI） is described after a brief review of some traditional and signal processing based methods. The pressure signal is first acquired using a microphone array, and used to reconstruct the 3-D sound field with some reconstruction algorithms. The reconstructed sound field is then shown graphically so that the main noise sources can be ＂seen＂ with eyes. The techniques are categorized in terms of noise signal acquiring methods, reconstruction algorithms, and display modes. Three commonly used reconstruction methods, the Fourier methods, Helmholtz-Kirchhoff equation based methods, and the equivalent source method, are investigated and compared in detail. Focus is given to the last method with some currently existing difficulties presented. Finally, a solution to the visualization application for NSI is proposed.[著者文摘]

Key words:noise source identification ; Fourier method ; Helmhohz-Kirchhoff equation method;equivalent sources method