臺灣博碩士論文加值系統

現今結構物健康監測技術多經由普通接觸式量測計（如：加速度規、速度規…等）裝設於結構物量取反應訊號進行分析判斷。為獲取結構細部反應資訊而大量安裝儀器的情況下，此種量測法將會有佈線繁雜以及裝設位置選項過少等問題產生。然而受惠於光學科技之進步，此問題可有效解決。光學量測方法是將待量測位置標示上一可經由影像辨識之目標點，藉由攝影裝置觀測目標移動情況進行三維動態分析。只要目標位於裝置監控範圍內，此方法將能進行大規模位置點計算。
本研究重點在探討此光學量測位移訊號應用於結構系統識別及破壞診斷的適用性。分析方法分為兩大類。第一類是整體系統識別方法，對於1.只需結構反應的斜方差型隨機子空間識別 (Covariance driven Stochastic Subspace Identification, SSI-COV)，和2.需系統輸入輸出資訊的子空間識別 (Subspace Identification, SI) 作探討，應用光學多維訊號進行自然頻率與阻尼比分析。再來是研究3.主成分分析 (Principal component analysis, PCA) 應用此訊號進行結構模態識別。第二類是局部系統評估，在將光學量測點網格化為數個單元後，應用幾何分析概念進行4.奇異譜分析 (Singular spectral analysis, SSA) 獲取單元之主要動態做進一步幾何處理。另外應用5.連續小波轉換 (Continuous Wavelet Transform, CWT) 分析訊號不連續性，對破裂做動時間點進行判斷。單元也可由6.有限元素法 (Finite element method, FEM) 計算其應變動態行為。本研究將會針對單層雙垮鋼筋混凝土桁架的振動台實驗進行實際應用。此實驗使用集成式光學量測儀器DMM (Dynamic Measuring Machine) 量測中間柱三維位移訊號。分析結果顯示應用此種空間位移訊號，整體系統識別方法可有效的獲取結構物資訊，並對結構變化進行描述；網格化之動態行為分析也能為結構局部破壞提供重要資訊。

In this research, the capability of advance spatial displacement measurement for structural health monitoring (SHM) is studied. The method for obtaining this kind of data is different from regular measuring system. It utilizes the optical processing technique to calculate the specific particles’ locations (called targets) within an image. While taking image and compute the locations over time, the dynamic motion can be estimated. This research employed the three dimensional displacement from optical sensors to identify system and perform damage assessment.
The applied signal analysis methodologies can separate into two categories, global system identification and local element motion detection. For global system, two subspace methods including 1.covariance-driven stochastic subspace identification (SSI-COV) and 2.recursive subspace identification (RSI) are examined. They can obtain the system natural frequency and damping ratio based on different condition. The other method is 3.principal component analysis (PCA), which the system normal modes can be briefly calculated while the measured locations are distributed along the system. For local motion, we can discretize the targets into a set of local elements. These elements motion is detect by 4.singular spectral analysis (SSA), 5.continuous wavelet transform (CWT), and 6.finite element method (FEM). The extracted information is used to describe the structural local properties and detect the damage occurrence. To examine the applications of these methodologies on real three dimensional displacement data, a shake table test of one-story two-bay RC frame performed in the NCREE is selected. This experiment installed a totally integrated optical measuring system (DMM, by NDI Inc.) on its central column to obtain the displacement. The analysis results show that this kind of data is capable for system identification, and the detection of damage is also feasible. Detail analyzes the discrete elements. The damage location may be obtained.

Authorization of Oral Members for Research Dissertation i
Acknowledgement iii
Abstract (in Chinese) v
Abstract vii
Contents ix
Table List xii
Figure List xiii

Chapter 1. Introduction 1
1.1 Motivation 1
1.2 Literature Review 2
1.3 Research Objective 4
Chapter 2. Signal Analysis Methodology 7
2.1 Introduction 7
2.2 Global System Characteristics Identification 7
2.2.1 Covariance-Driven Stochastic Subspace Identification (SSI-COV) 8
2.2.2 Recursive Subspace Identification (RSI) 12
2.2.3 Principal Component Analysis (PCA) 16
2.3 Local Element Motion Analysis 19
2.3.1 Singular Spectral Analysis (SSA) 20
2.3.2 Continuous Wavelet Transform (CWT) 23
2.3.3 Finite Element Method (FEM) 25
2.4 Chapter Summary 29
Chapter 3. Experimental Survey 33
3.1 Description of the Experiment 33
3.2 Preview of System Physical Properties 35
3.3 Optical Data Preprocessing 36
3.3.1 Three Dimensional Affine Transformation 36
3.3.2 Shifting of Target Positions 38
3.4 Global System Characteristics Identification 39
3.4.1 Global System Identification by SSI-COV 39
3.4.2 Global System Identification by RSI 41
3.4.3 Effective Mode Shape by PCA 43
3.4.4 Vector Space Damage Indicator by SSI-COV 44
3.5 Local Element Motion Analysis 47
3.5.1 Local element principal motion by SSA 47
3.5.2 Displacement non-continuity by CWT 51
3.5.3 Local element strain by FEM 52
3.5.4 Connection between Local Analysis Methodologies 54
3.6 Chapter Summary 55
Chapter 4. Conclusions 61
4.1 Research Conclusions 61
4.2 Recommendations for Future Work 65
References 66

1.Basseville, M., M. Abdelghani, and A. Benveniste, Subspace-based fault detection algorithms for vibration monitoring. Automatica, 2000. 36(1): p. 101-109.
2.Chang, C.-C., et al., Multi-component signal decomposition techniques for structural health monitoring. Proc. SPIE, 2005. 5765: p. 873.
3.Chang, C.-C., K.W. Sze, and Z. Sun, Structural damage assessment using principal component analysis. Proc. SPIE, 2004. 5394: p. 438-445.
4.Chao, S.-H. and C.-H. Loh, Application of SVD Techniques to Structural Damage Detection. Structural Health Monitoring, 2012.
5.De Boe, P. and J.-C. Golinval, Principal Component Analysis of a Piezosensor Array for Damage Localization. Structural Health Monitoring, 2003. 2(2): p. 137-144.
6.Elsner, J.B. and A.A. Tsonis, Singular Spectrum Analysis: A New Tool in Time Series Analysis1996: Plenum Press.
7.Feeny, B.F., ON PROPER ORTHOGONAL CO-ORDINATES AS INDICATORS OF MODAL ACTIVITY. Journal of Sound and Vibration, 2002. 255(5): p. 805-817.
8.Fish, J. and T. Belytschko, A First Course in Finite Elements2007: Wiley.
9.Golyandina, N., V.V. Nekrutkin, and A.A. Zhigljavsky, Analysis of Time Series Structure: SSA and Related Techniques2001: Chapman and Hall/CRC.
10.Han, S. and B.F. Feeny, Enhanced Proper Orthogonal Decomposition for the Modal Analysis of Homogeneous Structures. Journal of Vibration and Control, 2002. 8(1): p. 19-40.
11.Hattori, K., et al., Singular spectral analysis and principal component analysis for signal discrimination of ULF geomagnetic data associated with 2000 Izu Island Earthquake Swarm. Physics and Chemistry of the Earth, Parts A/B/C, 2006. 31(4–9): p. 281-291.
12.Huang, J.-R., System Identification of Degrading Hysteretic Restoring Forces of Reinforced Concrete Frames, in Department of Civil Engineering,2009, National Taiwan University.
13.Jolliffe, I.T., Principal Component Analysis. Vol. 2nd. 2002: Springer.
14.Li, J.-H., Enhancing Structural Seismic Response Insight of RC Frame by Time-Frequency Decomposition, in Department of Civil Engineering,2011, National Taiwan University.
15.Liu, Y.-C., Application of Covariance Driven Stochastic Subspace Identification Method, in Department of Civil Engineering,2011, National Taiwan University.
16.Mao, C.-H., Nonlinear System Identification Method for Structural Health Monitoring: Techniques for the Detection of Nonlinear Indicators, in Department of Civil Engineering,2009, National Taiwan University.
17.Moskvina, V. and A. Zhigljavsky, An Algorithm Based on Singular Spectrum Analysis for Change-Point Detection. Communications in Statistics - Simulation and Computation, 2003. 32(2): p. 319-352.
18.Pearson, K., LIII. On lines and planes of closest fit to systems of points in space. Philosophical Magazine Series 6, 1901. 2(11): p. 559-572.
19.Pelosi, G., The finite-element method, Part I: R. L. Courant [Historical Corner]. Antennas and Propagation Magazine, IEEE, 2007. 49(2): p. 180-182.
20.Takens, F., Dynamical systems and turbulence Warwick 1980. 1981. 898: p. 366-381.
21.Tselentis, G.A. and P. Paraskevopoulos, Site response analysis of Vartholomio W-Greece from singular spectrum analysis of microtremor and weak motion data. Soil Dynamics and Earthquake Engineering, 2010. 30(5): p. 378-394.
22.Van Overschee, P. and B.L.R. De Moor, Subspace Identification for Linear Systems: Theory - Implementation - Applications1996, Boston / London / Dordrecht: Kluwer Academic Publishers.
23.Vanlanduit, S., et al., A robust singular value decomposition for damage detection under changing operating conditions and structural uncertainties. Journal of Sound and Vibration, 2005. 284(3–5): p. 1033-1050.
24.Vautard, R., P. Yiou, and M. Ghil, Singular-spectrum analysis: A toolkit for short, noisy chaotic signals. Physica D: Nonlinear Phenomena, 1992. 58(1–4): p. 95-126.
25.Verhaegen, M., Identification of the deterministic part of MIMO state space models given in innovations form from input-output data. Automatica, 1994. 30(1): p. 61-74.
26.Weng, J.-H., Application of Subspace Identification in System Identification and Structural Damage Detection, in Department of Civil Engineering,2010, National Taiwan University.
27.Weng, J.-H. and C.-H. Loh, Recursive subspace identification for on-line tracking of structural modal parameter. Mechanical Systems and Signal Processing, 2011. 25(8): p. 2923-2937.
28.Zhang, Y., et al., Modal parameter identification using response data only. Journal of Sound and Vibration, 2005. 282(1–2): p. 367-380