天堂国产午夜亚洲专区-少妇人妻综合久久蜜臀-国产成人户外露出视频在线-国产91传媒一区二区三区

當前位置:主頁 > 科技論文 > 鑄造論文 >

高分辨率超聲Lamb波頻散曲線測量與板厚估計

發(fā)布時間:2018-05-09 03:26

  本文選題:超聲Lamb波 + 經典譜估計法; 參考:《數據采集與處理》2017年02期


【摘要】:超聲Lamb波在工業(yè)無損檢測與評價中具有廣闊的應用前景。但受制于超聲導波多模式頻散與混疊的影響,如何實現高分辨率超聲導波的模式分離與頻散曲線提取,進而定量分析材料的健康狀況一直是研究的熱點與難點。本文旨在將現代信號處理中廣泛采用的譜估計技術應用于超聲Lamb波陣列信號分析,并通過提取頻散曲線定量測量板厚度。以經典參數譜估計法中的Yule-Walker法與Burg法為例,實驗測量和分析了3mm,4mm和5mm厚鋁板中的寬帶多模式Lamb波信號,準確地提取了鋁板中的寬帶Lamb波頻散曲線,并比較了兩種經典譜估計方法與二維傅氏變換法的性能,最終實現了鋁板厚度估計。
[Abstract]:Ultrasonic Lamb wave has a broad application prospect in industrial nondestructive testing and evaluation. However, due to the influence of multi-mode dispersion and aliasing of ultrasonic guided waves, how to realize the mode separation and dispersion curve extraction of high-resolution ultrasonic guided waves and to quantitatively analyze the health status of materials has always been a hot and difficult issue. In this paper, the widely used spectral estimation technique in modern signal processing is applied to the analysis of ultrasonic Lamb wave array signals, and the dispersion curve is extracted to quantitatively measure the thickness of the plate. Taking the Yule-Walker method and Burg method of classical parameter spectrum estimation as examples, the broadband multi-mode Lamb wave signals in 3mm 4mm and 5mm thick aluminum plates are measured and analyzed experimentally, and the dispersion curves of broadband Lamb waves in aluminum plates are accurately extracted. The performance of two classical spectrum estimation methods and two dimensional Fourier transform method are compared and the thickness estimation of aluminum plate is realized.
【作者單位】: 巴黎第六大學法國國家科學研究院國家衛(wèi)生與醫(yī)學研究院生物醫(yī)學成像實驗室(LIB);復旦大學電子工程系;上海市醫(yī)學圖像處理與計算機輔助手術重點實驗室;
【基金】:國家自然科學基金(11304043,11327405,11525416)資助項目
【分類號】:TG115.285

【相似文獻】

相關期刊論文 前1條

1 徐江;付曉明;武新軍;康宜華;;基于檢測信號的導波頻散曲線獲取方法[J];無損檢測;2011年10期

,

本文編號:1864370

資料下載
論文發(fā)表

本文鏈接:http://www.sikaile.net/kejilunwen/jiagonggongyi/1864370.html


Copyright(c)文論論文網All Rights Reserved | 網站地圖 |

版權申明:資料由用戶2c347***提供,本站僅收錄摘要或目錄,作者需要刪除請E-mail郵箱bigeng88@qq.com