基於濾除環境效應後振態頻率變化之辦公大樓損傷偵測

利用本團隊新近發展的隨機子空間識別機制，本研究針對典型辦公大樓建立一套迅速有效的損傷偵測工具。首先從中科管理局及工商大樓21個月的長期微振量測進行各項振態參數的變化趨勢探討，然後取頂樓加速度訊號的標準差來代表結構之均方根反應，據以充分反映主要由風速和大樓工作人員活載之消長所決定的外力擾動強度，結果發現其正是影響各振態參數的最直接環境因素。接著針對目標振態頻率統計其與均方根加速度的關係，從而擬合出最佳化指數函數，並據以排除環境因素的影響。最後則利用濾除環境效應後之振態頻率變化，取其單日移動平均與移動標準差，經由統計分析發展出兩項損傷指標來建立敏感的樓房損傷偵測機制。

關鍵字

隨機子空間識別；損傷偵測；環境效應；振態頻率變化

並列摘要

A stochastic subspace identification algorithm was recently developed by this research group for accurately identifying the modal parameters. Based on such a new method, the current study establishes an effective methodology for the damage detection of typical office building structures. The modal frequencies, damping ratios, and shape vectors are first identified from the ambient vibration measurements of CTSPB Building and IC Building. The variation trends for all the modal parameters are then investigated according to 21 months of long-term measurements, followed by the assessment of influences from different environmental factors. The standard deviation of the top floor acceleration measurement is taken in this study to represent the root-mean-square response of the structure and is observed to be the most dominant environmental factor. It is found that all modal frequencies are highly correlated with the root-mean-square acceleration in a negative manner. The relationship between the target frequency and the average acceleration response is subsequently analyzed to obtain an optimal exponential function for excluding the environmental effects. Finally, the daily moving average and standard variation for the long-term variation of the modal frequency residue are taken to develop two damage indices for establishing a sensitive building damage detection algorithm.

並列關鍵字

stochastic subspace identification ； damage detection ； environmental effects ； variation of the modal frequencies

主題瀏覽