應用量測訊號之時頻分析技術於結構健康診斷

Time-frequency analysis techniques using measurement signals for Structural Health Monitoring

李全富(Chuan-Fu Li)

指導教授 ： 羅俊雄

國立臺灣大學/工學院/土木工程學研究所/碩士(2018年)

https://doi.org/10.6342/NTU201801317

摘要

在結構健康診斷中，時頻分析方法因其具有時間頻率的定位功能其已被廣泛應用於取得結構振動特徵。本研究分別討論基於連續小波轉換、連續小波轉換改變小波中心頻率、改良式小波包轉換、同步擠壓小波轉換、平滑重排偽維格納分布(RSPWVD)，這五種不同的時頻分析方法的結構康檢測方法其中包含(1) Novelty Index(2) 以小波為基礎的似然比檢驗(3) Time varying damage index(4) 使用平滑重排偽維格納分佈識別地震反應結構物頻率的變化(5) instantaneous softening並介紹基於同步擠壓小波轉換的模態識別方法。本研究使用兩組振動台試驗包含八層樓剪力鋼構架以及鋼筋混凝土試驗以及三棟實際結構物包含中興大學土木環工大樓、交大公教大樓以及嘉南大圳白河橋驗證上述方法的可行性，並討論不同案例的結果。最後本文藉由同步擠壓轉換可以重建訊號的特性，重建出近斷層地震脈衝似歷時並使用sap2000非線性歷時分析進行模擬並藉由模擬結果的得到脈衝似歷時特性。

關鍵字

結構健康診斷 ； 時頻分析方法 ； 同步擠壓小波轉換 ； 平滑重排偽維格納 分布 ； 結構損傷評估 ； 結構損傷部位識別

並列摘要

In the structural health monitoring, the time-frequency analysis method has been widely used to obtain structural vibration characteristics because of its time-frequency localization property. This study discusses the structural health detection methods based on five different time-frequency analysis methods: Continuous wavelet transform, Continuous wavelet transform with variable central frequency, Enhanced wavelet packet transform, Synchrosqueezed wavelet transform, and Reassigned smoothed pseudo Wigner-Ville distribution. Including (1) Novelty Index(2) Wavelet-based likelihood ratio test (3) Time varying damage index (4) Use Reassigned Smoothed pseudo Wigner-Ville distribution to identify changes in structural frequency (5) instantaneous softening and introducing Synchrosqueeze-based Wavelet Transform method to identify modal parameter. This study used two sets of shaking table tests including the eight-story shear steel frame and the reinforced concrete test and the three actual structures including the NCHU Civil & Environmental Engineering Building, NCTU Government Employee Housing and Bai-Ho Bridge to verify the above method. Finally ,this article will use Synchrosqueezed wavelet transform to reconstruct pulse like wave and use Stick Model Nonlinear History Analysis to get Pulse like wave characteristics.

並列關鍵字

structural health monitoring ； time-frequency analysis method ； Synchrosqueezed wavelet transform ； Reassigned smoothed pseudo Wigner-Ville distribution ； damage detection ； damage localization

參考文獻

[1] S.T. Quek, Q. Wang, L. Zhang, K.-K. Ang, Sensitivity analysis of crack detection in beams by wavelet technique, International Journal of Mechanical Sciences 43 (2001) 2899–2910.

[2] A. Klepka, T. Uhl, Identification of modal parameters of non-stationary systems with the use of wavelet based adaptive filtering, Mech. Syst. Signal Process. 47 (2014) 21–34.

[3] A. Tjirkallis, and A. Kyprianou, Damage detection under varying environmental and operational conditions using Wavelet Transform Modulus Maxima decay lines similarity. Mechanical Systems & Signal Processing s 66–67(2016) 282-297.

[4] S. Banerji, T.B Roy, A. Sabamehr, A. Bagchi. Experimental validation of a structural damage detection method based on marginal Hilbert spectrum. SPIE Smart Structures and Materials Nondestructive Evaluation and Health Monitoring, 2017, Portland, Oregon, United States. (2017)

[5] R. Tarinejad, M. Damadipour, Modal identification of structures by a novel approach based on FDD-wavelet method, Journal of Sound and Vibration 333 (2014) 1024–1045.