| 摘要: | 隨著人類文明的高度發展與進步，為了追求方便、舒適與迅速的生活，人造的結構物也越顯得大型與精緻。為確保結構物的可靠性與安全性，可藉由對結構體的動態特性，求解出結構系統的參數，作為設計考量、振動控制與利用振動訊號進行監測和結構診斷的依據。在眾多領域中，已經發展了許多的系統識別法，而各個的識別方法都有其優點與缺點，最重要的是如何選擇最合適的系統識別方法。安裝加速度計、速度計等量測振動訊號儀器於結構物上，量測結構物的微小振動訊號，將此訊號放入系統識別的程式軟體中，搭配訊號分析相關知識與結構物的力學計算，與數值模擬實驗的解析解進行判斷與解析，即可計算出結構物的相關振動特性。使用頻譜分析、時域分析與子空間分析等進階分析方法，探討量測結構物的健康管理與相關診斷。
本論文探討透過隨機遞減法與多重隨機遞減法萃取出第一模態的模態頻率與模態振形，並使用頻率域分解法與隨機子空間識別，為僅需要輸出反應的模態參數識別方法，在無使用隨機遞減法、使用隨機遞減法與使用多重隨機遞減法時，是否會影響識別各個模態的模態頻率與模態振形。並且評估這些方法在土木結構物上，識別模態振形的可行性。搭配加快速傅立葉轉換，與上述兩種識別方法進行自然頻率的識別比較。
在本文章中，使用有限元模型進行模擬，施加外在的高斯白噪訊於試驗梁上，量測到的梁微小振動反應進行以上方式分析，並驗證分析方法的試用性。最後，將這些方法應用於現地量測中，以求取橋梁的模態振形以及自然頻率，並與本團隊製作的現地橋梁之模擬模型進行比對。;With the highly developed and progressive civilization, artificial structures have become increasingly large-scale and sophisticated in pursuit of convenience, comfort, and rapid living.
To ensure the reliability and safety of these structures, their dynamic characteristics can be analyzed to derive system parameters, which serve as a basis for design considerations, vibration control, and the use of vibration signals for monitoring and structural diagnosis. In various fields, numerous system identification methods have been developed, each with its own strengths and weaknesses. The most crucial aspect is how to select the most appropriate system identification method. By installing instruments such as accelerometers and velocity sensors on structures, the small vibration signals of the structures can be measured. These signals are then input into system identification software, along with relevant signal analysis knowledge and
structural mechanics calculations, and combined with numerical simulation experiments and analytical solutions. This process allows for the calculation of the structural vibration properties.
Advanced analysis methods, such as spectrum analysis, time-domain analysis, and subspace analysis, can be used to explore the health management and related diagnosis of measured structures.
This thesis investigates the extraction of the first mode′s modal frequency and modal shape using Random Decrement (RD) and Multiple Random Decrement (MRD) methods. It employs
Frequency Domain Decomposition (FDD) and Random Subspace Identification (SSI) as modal parameter identification techniques that only require output responses.The study aims to assess whether the utilization of RD, RD combined with MRD, or solely MRD affects the identification of modal frequencies and modal shapes for various modes. Additionally, it evaluates the feasibility of these methods in identifying modal shapes for civil engineering structures. Fast Fourier Transform (FFT) is utilized to accelerate the frequency identification process and compared with the aforementioned identification methods.
In this paper, finite element models are employed for simulation, and external Gaussian white noise is applied to a test beam. The small vibration responses of the beam obtained from measurements are analyzed using the methods mentioned above, and the applicability of these analysis techniques is validated. Finally, these methods are applied to on-site measurements to obtain the modal shapes and natural frequencies of bridges and compared with simulated
models of bridges produced by our team. |
