Skip navigation

DSpace

機構典藏 DSpace 系統致力於保存各式數位資料(如:文字、圖片、PDF)並使其易於取用。

點此認識 DSpace
DSpace logo
English
中文
  • 瀏覽論文
    • 校院系所
    • 出版年
    • 作者
    • 標題
    • 關鍵字
  • 搜尋 TDR
  • 授權 Q&A
    • 我的頁面
    • 接受 E-mail 通知
    • 編輯個人資料
  1. NTU Theses and Dissertations Repository
  2. 工學院
  3. 應用力學研究所
請用此 Handle URI 來引用此文件: http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/92107
完整後設資料紀錄
DC 欄位值語言
dc.contributor.advisor劉佩玲zh_TW
dc.contributor.advisorPei-Ling Liuen
dc.contributor.author曾勁凱zh_TW
dc.contributor.authorJing-Kai Tsengen
dc.date.accessioned2024-03-05T16:19:18Z-
dc.date.available2024-03-06-
dc.date.copyright2024-03-05-
dc.date.issued2024-
dc.date.submitted2024-02-17-
dc.identifier.citationSansalone, M. C., N. J., 1986, Impact-echo : a method for flaw detection in concrete using transient stress waves, U.S. Dept. of Commerce, National Institute of Standards and Technology, Gaithersburg, MD.
Lin, Y. C., and Sansalone, M., 1992, "Transient-response of thick circular and square bars subjected to transverse elastic impact," J. Acous. Soc. Am., 91(2), pp. 885-893.
Lin, Y. C., and Sansalone, M., 1992, "Detecting flaws in concrete beams and columns using the impact-echo method," ACI Materials Journal, 89(4), pp. 394-405.
Lin, Y. C., and Su, W. C., 1996, "Use of stress waves for determining the depth of surface-opening cracks in concrete structures," ACI Materials Journal, 93(5), pp. 494-505.
Wang, C. Y., Chiu, C. L., Tsai, K. Y., Chen, P. K., Peng, P. C., and Wang, H. L., 2014, "Inspecting the current thickness of a refractory wall inside an operational blast furnace using the impact echo method," NDT & E International, 66, pp. 43-51.
Xu, J. M., Wang, P., An, B. Y., Ma, X. C., and Chen, R., 2018, "Damage detection of ballastless railway tracks by the impact-echo method," Proceedings of the Institution of Civil Engineers-Transport, 171(2), pp. 106-114.
Yao, F., Chen, G. Y., and Abula, A., 2018, "Research on signal processing of segment-grout defect in tunnel based on impact-echo method," Construction and Building Materials, 187, pp. 280-289.
Yao, F., and Chen, G. Y., 2019, "Time-frequency analysis of impact echo signals of grouting defects in tunnels," Russian Journal of Nondestructive Testing, 55(8), pp. 581-595.
Liu, P. L., Tsai, C. D., and Wu, T. T., 1996, "Imaging of surface-braking concrete cracks using transient elastic waves," NDT & E International, 29(5), pp. 323-331.
Ohtsu, M., and Watanabe, T., 2002, "Stack imaging of spectral amplitudes based on impact-echo for flaw detection," NDT & E International, 35(3), pp. 189-196.
劉佩玲, 葉柏涼, and 游志源, "Imaging of concrete defects using elastic wave tests 混凝土缺陷影像法," pp. 41-50.
Tong, J. H., Liao, S. T., and Lin, C. C., 2007, "A new elastic-wave-based imaging method for scanning the defects inside the structure," IEEE Transactions on Ultrasonics Ferroelectrics and Frequency Control, 54(1), pp. 128-137.
Schubert, F., Wiggenhauser, H., and Lausch, R., 2004, "On the accuracy of thickness measurements in impact-echo testing of finite concrete specimens-numerical and experimental results," Ultrasonics, 42(1-9), pp. 897-901.
Liu, P. L., and Yeh, P. L., 2010, "Vertical spectral tomography of concrete structures based on impact echo depth spectra," NDT & E International, 43(1), pp. 45-53.
Pratt, D., and Sansalone, M., 1992, "Impact-echo signal interpretation using artificial-intelligence," ACI Materials Journal, 89(2), pp. 178-187.
Zhang, J. K., Yan, W. Z., and Cui, D. M., 2016, "Concrete condition assessment using impact-echo method and extreme learning machines," Sensors-Basel, 16(4).
林昀儒, 2020, "以CNN自動編碼器辨識敲擊回音試驗之異常訊號," 碩士, 國立臺灣大學, 台北市.
陳柏合, 2021, "以深度學習判別鋼筋與裂縫之敲擊回音雙譜," 碩士, 國立臺灣大學, 台北市.
陳源泰, 2021, "以深度學習方法判讀敲擊回音時頻圖," 碩士, 國立臺灣大學, 台北市.
謝承展, 2022, "以機器學習與主成分分析進行敲擊回音本質模態函數之分類," 碩士, 國立臺灣大學, 台北市.
Achenbach, J. D., 1973, Wave propagation in elastic solids / by J. D. Achenbach, North-Holland Pub. Co., Amsterdam.
Goldsmith, W., 1960, Impact : the theory and physical behaviour of colliding solids, Edward Arnold (Publishing) Ltd, London.
Colla, C., and Lausch, R., 2003, "Influence of source frequency on impact-echo data quality for testing concrete structures," NDT & E International, 36(4), pp. 203-213.
Lin, Y., Sansalone, M., and Carino, N. J., 1990, "Finite element studies of the impact-echo response of plates containing thin layers and voids," Journal of Nondestructive Evaluation, 9(1), pp. 27-47.
-
dc.identifier.urihttp://tdr.lib.ntu.edu.tw/jspui/handle/123456789/92107-
dc.description.abstract敲擊回音法為廣泛應用於混凝土結構之非破壞檢測技術,主要是藉由外力敲擊待測物產生應力波後,將結構物表面反應的時間域訊號轉換成頻率域訊號,並根據頻譜中的尖峰頻率計算出潛在缺陷之對應深度。當量測範圍擴大時,深度頻譜斷層掃瞄法不僅能夠同時對比測線上的多筆訊號,也能更直觀地了解結構內部情況。然而,敲擊回音訊號通常會摻雜雜訊以及來自邊界的反射訊號,訊號頻寬也會受敲擊源影響,使深度頻譜斷層掃瞄影像出現雜訊、干擾與暗帶,即使具經驗的檢測人員在判讀影像時都會遭遇困難。
本研究嘗試發展一套能夠使斷層掃瞄影像中的裂縫區域更明顯之深度學習模型,並由輸出影像進一步標記出裂縫位置、尺寸及走向。本研究用以強化影像的深度學習模型為卷積自動編碼器(Convolutional Autoencoder, CAE),CAE憑藉著非監督式學習之演算法特性,對於影像的異常檢測具有良好的效果。本研究先以有限元素分析軟體模擬無裂縫之半無限域,將受6mm鋼珠敲擊之回音訊號繪製斷層掃瞄影像後用以訓練CAE模型,使該模型具備還原無裂縫斷層掃瞄影像之能力。將內含裂縫之斷層掃瞄影像輸入訓練完成之CAE模型後,該模型便會盡可能地還原無裂縫區域的影像。將輸入影像扣除輸出之重構影像所得到的殘差影像便可清楚顯示裂縫區域。接著,本研究先描繪出殘差影像中各亮帶之邊界,再由各亮帶範圍內之像素亮度加總,以選出最可能的裂縫區域,再透過加權線性迴歸以一條線段標示出裂縫位置及走向。
為提升CAE模型之實用性與偵測裂縫之精度,本研究分別對訓練影像添加不同大小的雜訊、選擇不同激勵函數及改變CAE模型架構深度訓練模型,並以殘差影像之背景/缺陷亮度比做為評比指標,以找出較優的CAE模型。經過數值模擬影像測試,當編碼器和解碼器各為3層,以整流線性單位(Rectified Linear Unit, ReLU)函數作為激勵函數,並以無雜訊及添加3%雜訊之影像進行訓練而成的CAE模型表現最佳。
在標記裂縫影像部分,本研究分別模擬不同尺寸、深度及走向之裂縫斷層掃瞄影像。大多數裂縫影像所標記之裂縫深度誤差皆在1cm以內,在水平跨度誤差則大部分在5cm以內,惟有深度6cm和16cm之水平裂縫的水平跨度誤差較為不規律。在標記實驗裂縫影像部分,深度10cm之水平裂縫的水平跨度為32cm,迴歸標記的裂縫水平跨度為26cm,誤差為6.0cm,深度誤差為0.1cm;深度12~8cm之傾斜裂縫的水平跨度亦為32cm,迴歸標記的裂縫水平跨度為36cm,誤差為4.0cm,深度誤差則為0.2cm。
為測試模型的泛用性,本研究嘗試以不同大小鋼珠作為敲擊源進行試驗。測試結果發現當鋼珠直徑不為6mm時,只需將斷層掃瞄影像之強度及垂直軸以鋼珠直徑等比例調整,CAE模型便能有效地消除不必要雜訊,將斷層掃瞄影像中的裂縫區域明顯標示出來。
綜上所述,本研究訓練之CAE模型能夠有效消除敲擊回音斷層掃瞄影像中的背景訊號及雜訊干擾並凸顯裂縫,而線性迴歸模型也能大致偵測並標記出裂縫的位置及走向,可作為檢測人員判讀混凝土內部情況的有效工具。
zh_TW
dc.description.abstractThe impact-echo method is a widely used nondestructive testing technique for concrete structures. It involves generating stress waves in the test object by applying an impact force and converting the resulting surface response from the time domain into the frequency domain. The depth of a potential defect is then calculated based on the peak frequency in the spectrum. By conducting a series of impact-echo tests along a test line, one may use depth spectral tomography to produce a more intuitive view of the internal structure. However, depth spectral tomogram often contain noise, interferences, and dark zones. Even experienced inspectors may have difficulty interpreting the images.
This study aims to develop a methodology to enhance the crack image in the impact-echo depth spectral tomograms. A convolutional autoencoder (CAE) is adopted to this end because it can detect anomalies in images.The finite element analysis has been used to simulate the response of a perfect half-space due to the impact of a 6mm steel ball. By training the CAE with simulated tomograms, the model learns their features and is able to reconstruct them. As the tomogram of the specimen that contains cracks is input into the CAE, the model tries to reproduce the crack-free image. After subtracting the reconstructed tomogram from the input tomogram, the unwanted noise, interferences, and dark zones are diminished in the residual tomogram, but the crack image remains intact and noticeable. Apply edge detection to find the bright zones in the residual tomogram. Then, use the weighted linear regression to draw a line segment that denotes the crack.
The influence of several factors on the performance of CAE models has been studied, including the noise level in the training tomograms, choice of activation function, and depth of the CAE model architecture. The CAE models are compared based on the residual images' background-to-defect brightness ratio. The CAE model with three layers for both encoder and decoder, using rectified linear unit (ReLU) as the activation function, trained with noise-free and 3% noise-added tomograms, yields the best results.
Labeling cracks by weighted linear regression has been conducted on the residual tomograms of specimens with various crack sizes. For the simulated tomograms, depth errors are primarily within 1cm, and the errors in crack horizontal spans are generally within 5cm for 32cm span cracks. For the experimental tomograms, labeling of a 10cm depth, 32cm span horizontal crack exhibits depth and span errors of 0.1cm and 6.0cm, respectively; labeling of a 32cm span inclined cracks with depth varying from 8 to 12 exhibits depth and span errors of 0.2cm and 4.0cm, respectively.
To examine the versatility of the CAE model, tomograms generated by various impact balls have been tested. It is found that the results are not satisfactory as the diameter of the steel ball changes. However, if the intensity and vertical scale of the tomograms are scaled according to the diameter, the model can effectively enhance the crack images in the residual tomograms.
In summary, this study's CAE model can effectively eliminate unwanted noise, interference, and dark zones from the impact-echo depth spectral tomogram. As such, it highlights the crack image. The weighted linear regression can roughly label the cracks in the residual tomogramss. It can assist inspectors in interpreting the internal conditions of concrete structures.
en
dc.description.provenanceSubmitted by admin ntu (admin@lib.ntu.edu.tw) on 2024-03-05T16:19:18Z
No. of bitstreams: 0
en
dc.description.provenanceMade available in DSpace on 2024-03-05T16:19:18Z (GMT). No. of bitstreams: 0en
dc.description.tableofcontents口試委員會審定書 I
誌謝 II
摘要 III
Abstract V
目次 VII
圖次 IX
表次 XIII
第一章 前言 1
1-1 研究背景 1
1-2 文獻回顧 1
1-3 研究動機與目的 3
1-4 全文簡介 4
第二章 敲擊回音深度頻譜斷層掃瞄 6
2-1 敲擊回音法 6
2-2 敲擊回音主要參數 9
2-2-1 敲擊源 9
2-2-2 取樣時間 10
2-2-3 取樣時距 11
2-3 深度頻譜斷層掃瞄法 12
第三章 卷積自動編碼器 16
3-1 模型架構 17
3-2 裂縫標記步驟 22
3-3 模型評比指標 26
第四章 資料集 27
4-1 數值模擬 27
4-1-1 數值模擬步驟 27
4-1-2 模擬試體 28
4-2 模型試驗 39
4-2-1 實驗設備 39
4-2-2 實驗試體 39
4-3 訓練集與測試集 46
第五章 結果與討論 50
5-1 不同模型配置結果比較 50
5-1-1 時間域訊號添加雜訊之影響 52
5-1-2 不同激勵函數之影響 61
5-1-3 模型架構深淺之影響 64
5-2 裂縫標記結果 68
5-3 模型泛用性測試 87
5-3-1 試體厚度未知之斷層掃瞄影像 87
5-3-2 以不同大小鋼珠敲擊之訊號繪製斷層掃瞄影像 92
第六章 總結與未來展望 100
參考文獻 102
-
dc.language.isozh_TW-
dc.title以卷積自動編碼器進行敲擊回音深度頻譜斷層掃瞄之裂縫偵測zh_TW
dc.titleA Convolutional Autoencoder for Crack Detection on Impact-Echo Depth Spectral Tomogramsen
dc.typeThesis-
dc.date.schoolyear112-1-
dc.description.degree碩士-
dc.contributor.oralexamcommittee郭茂坤;孫嘉宏zh_TW
dc.contributor.oralexamcommitteeMao-Kuen Kuo;Jia-Hong Sunen
dc.subject.keyword敲擊回音法,非破壞檢測,深度頻譜斷層掃瞄,深度學習,卷積神經網路,自動編碼器,裂縫偵測,zh_TW
dc.subject.keywordimpact-echo method,nondestructive test,depth spectral tomogram,deep learning,convolution neural network,autoencoder,crack detection,en
dc.relation.page104-
dc.identifier.doi10.6342/NTU202400498-
dc.rights.note同意授權(全球公開)-
dc.date.accepted2024-02-17-
dc.contributor.author-college工學院-
dc.contributor.author-dept應用力學研究所-
顯示於系所單位:應用力學研究所

文件中的檔案:
檔案 大小格式 
ntu-112-1.pdf4.21 MBAdobe PDF檢視/開啟
顯示文件簡單紀錄


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。

社群連結
聯絡資訊
10617臺北市大安區羅斯福路四段1號
No.1 Sec.4, Roosevelt Rd., Taipei, Taiwan, R.O.C. 106
Tel: (02)33662353
Email: ntuetds@ntu.edu.tw
意見箱
相關連結
館藏目錄
國內圖書館整合查詢 MetaCat
臺大學術典藏 NTU Scholars
臺大圖書館數位典藏館
本站聲明
© NTU Library All Rights Reserved