應用小波分析與功率譜密度法於結構健康監測之研究

Application of Wavelet Analysis and Power Spectral Density on Structural Health Monitoring

問世賢(Shih-Hsian Wen)

指導教授 ： 洪士林

國立交通大學/工學院/土木工程系所/碩士(2009年)

https://doi.org/10.6842/NCTU.2009.00065

摘要

由於結構健康監測議題日益受到重視，多種結構系統識別方式如雨後春筍般提出，小波分析在近年來應用在結構系統識別方面有著相當程度的優越性，而功率譜密度法以其在頻率域較為敏感的特性，廣泛的應用在各種不同之訊號處理上。 本論文係分成三個部份來探討小波分析與功率譜密度法於結構健康監測之研究:先以數值模擬一簡支梁，藉由改變簡支梁部分鋼梁斷面以模擬損壞；而後使用國家地震研究中心五層樓鋼構之試驗資料，以加斜撐之方式改變構架性質；最後以一自製之小型三層樓鋼構架，由小型震動台做地震模擬，並以無線感應器接收構架各層之加速度變化。將以上三種訊號先利用小波分析的多重解析技術找出結構特性之變化，其後利用功率譜密度法比較構架未損壞與損壞後之頻率相位差異，建立一識別參數Di，觀察功率譜密度法是否可明確的指出構架損壞的位置，由結果顯示在適當之調整門檻值α與擷取頻率fcf之下，可以概括地找出構架之損害位置。

關鍵字

小波分析 ； 功率譜密度 ； 結構健康監測

並列摘要

Since the research of structural health monitoring (SHM) is paid more attention recently, many different approaches of structural system identification (SI) and damage detection have been developed. In which, wavelet analysis is one of successes approach in structural system identification. The power spectral density (PSD) has comparatively sensitive characteristic in frequency domain, It., hence, is employed in the problems of structural damage detection.
The objective of this research is applying wavelet analysis and PSD approaches to identify the structural properties, such as natural frequencies and damping ratios, and detect the location of damage, if it occurred, of structures. Three different cases were employed to verify the feasibility of proposed approaches. One is a numerical example of simple supported beam with different damaged cases. The other two are experiment examples, a five-story steel frame in NCREE and a small three-story test steel frame. The structural properties, natural frequencies and damping ratios, of three examples with different damages were identified via wavelet analysis. The results indicate that the approach of wavelet analysis is a sufficient method for SI. Thereafter, the PSD method were employed to identify the location of damage occurred via comparing the values of absolute difference PSD method (ADPSD) and curvature difference PSD method (CDPSD) between damaged and undamaged cases. By properly selecting threshold value α and cutoff frequency fcf, two proposed PSD approaches can roughly detect the location of damage occurred in which segment of beam or floor of frames.

並列關鍵字

wavelet analysis ； power spectral density ； structural health monitoring

參考文獻

[11] Quan Wang，Xiaomin Deng，“Damage detection with spatial wavelets”，International Journal of Solids and Structures，36，3443-3468，1999.

[1] Hoon Sohn，“Effects of Environmental and Operational Variability on Structural Health Monitoring”，Philosophical Transactions of the Royal Society，2007.

[5] H. Sohn，K. Worden，C.R. Farrar， “Statistical damage classification under changing environmental and operational conditions”，Journal of Intelligent Materials Systems and Structures，13(9)， 561-574，2003.

[6] G. Park，H. Cudney，D. Inman， “Impedance-based health monitoring technique for massive structures and high-temperature structures”，SPIE，3670，461-469，1999.

[12] Ser-Tong Quek， Quan Wang，Liang Zhang，Kian-Keong Ang，“Sensitivity analysis of crack detection in beams by wavelet technique”，International Journal of Mechanical Sciences，43，2899–2910，2001.