2015-11-012024-05-18https://scholars.lib.ntu.edu.tw/handle/123456789/710207摘要：基礎設施的性能維繫著國家整體的經濟發展與人民生命財產安全，基礎設施包括了各式的建築物、橋梁、交通系統及維生管線等，其中的任何設施都必須長時間保持原有的功能性，不然數以萬計的經濟損失與人員死傷將會發生，衝擊整體社會。為了預防災害的衝擊，結構健康監測藉由感測系統獲得結構相關資訊，經由損傷識別的方法，及早判別結構的惡化與損傷，預防且降低災害造成的損失。在此研究中，將建置一個全新的結構健康監測的平台，結合時頻域分析和矩陣分解法，能夠對量測訊號做相應的處理，並且透過結構反應估計和特徵提取的方法，實現結構損傷識別，完成此結構健康監測平台。對於量測訊號處理，將透過時頻域分析，有效降低時間域訊號的雜訊，使頻率域反應的相位平滑化，並消除結構量測反應中不正常的訊號。另外，本研究將開發結構反應估計的方法，利於判別結構損傷發生的時間，配合一系列的特徵提取方法，有效地判定結構損傷的位置與程度，達成預警與預防的效果。為了瞭解該平台的性能，本研究將考慮大量的參數分析，並配合數值模擬與振動台實驗驗證，確認此平台的可行性。最終此結構健康監測平台的目標是實現連續自主式的結構健康監測方法，以利及早發現結構性能的缺失。<br> Abstract: Performance of civil infrastructure directly affects public safety and society cost. Civil infrastructure refers to the integration of various systems such as buildings, bridges, transportation networks, lifeline, etc. Components in such systems need to be functional all the time; otherwise, a huge amount of economic loss and dead lives would occur and then impact the entire society. Structural health monitoring serves as a tool that enables structural information to be acquired through sensing technology, and is of need to early diagnose deterioration and damages in structures. Health monitoring strategies are often realized through a combination of high-quality sensing systems and high-performance structural integrity assessment methods; therefore, structural deviations can be effectively identified by interpreting the raw sensor measurements using signal processing techniques. The proposed research is to develop a structural health monitoring framework that applies matrix factorization techniques to the time-frequency representation of multi-channel measured signals. This method will process vibrational input and/or output responses of structures to improve raw data quality, to estimate structural responses, to derive signal features, and to detect structural variations. For example, the proposed framework will be capable of reducing noise in time-domain responses, smoothing phases for frequency-domain responses, removing abnormal signals due to electric effects. In addition, this framework allows estimating structural responses based on the relationship between the current and past measurements as well as deriving signal-based features from time-frequency distributions. By integrating the response prediction with feature-based damage detection, structural integrity can be assessed in terms of damage time occurrences, locations, quantifications. With all these developments, a complete signal-based structural health monitoring framework will be established which will benefit structural diagnosis and early warning to structural damages. Performance of the proposed framework will also be evaluated through a parametric study. To validate the proposed framework, both numerical and experimental tests will be conducted. The experimental work is also planned to verify the proposed framework for identifying changes of structural properties using shaking table testing as well as for exploring the feasibility in the structural health monitoring applications. The ultimate goal in this research is to continuously and autonomously implement the proposed framework in a long-term health monitoring system and then to early diagnose structural deficiency.