指導教授：羅俊雄臺灣大學：土木工程學研究所熊婉贏Hsiung, Wan-YingWan-YingHsiung2014-11-252018-07-092014-11-252018-07-092014http://ntur.lib.ntu.edu.tw//handle/246246/260867本研究主要目的是藉由實驗室試驗和現地試驗資料驗證小型旋轉風機葉片模態分析(Modal Analysis)、系統識別(System Identification)與損傷偵測(Damage Detection)。風機葉片上安裝了加速度感測器和應變計，透過無線感測系統來量測葉片振動反應，本研究的第一部分實驗進行於國家地震中心(National Center for Research on Earthquake Engineering, NCREE)，設計一台小型研究規模的風機系統，葉片為金屬材料製成，將感測器安裝於葉片上，利用馬達使葉片旋轉，風機架設於振動台上，在有輸入地震力作用時，風機旋轉葉片與機塔存在互制行為情況下，量測不同轉動速度的葉片振動訊號。第二部分為現地實驗於加州大學戴維斯分校(University of California, Davis)工學院大樓(Bainer Hall)頂樓小規模的水平軸風機系統，型號為Bergey Windpower XL.1，兩個實驗主要的差別為，現地實驗依賴真實風力作用，而不是利用馬達控制葉片的轉動。實驗室測試中，風機葉片主要分成兩種，未受到損壞葉片和損壞葉片，在連接處鬆脫兩顆螺絲稱之為損壞葉片，使用損壞偵測指標識別葉片是否有損傷發生。此外利用訊號處理和系統識別技術分析兩個實驗資料，識別葉片動態特性，分析方法包括：多點奇異譜分析法(Multivariate Singular Spectrum Analysis, MSSA)、協方差隨機子空間識別法(Covariance-driven Stochastic Subspace Identification, SSI-COV)、子空間識別法(Data-driven Subspace Identification Method, SI-DATA)。上述實驗皆可確認運行中風機葉片的動態特性，實驗結果期許作為往後風機監測研究基準。The objective of this study was to validate modal analysis, system identification and damage detection of small-scale rotating wind turbine blades in the laboratory and in the field. Here, wind turbine blades were instrumented with accelerometers and strain gages, and data acquisition was achieved using a prototype wireless sensing system. In the first portion of this study conducted in the laboratory, sensors were installed onto metallic structural elements that were fabricated to be representative of an actual wind blade. In order to control the excitation (rotation of the wind blade), a motor was used to spin the blades at controlled angular velocities. The wind turbine was installed on a shaking table for testing. Under the circumstance of the couple of rotating wind turbine blade and tower with base excitation to measure the signals of different rotation speed. On the other hand, the second part of this study which is field experiment utilized a small-scale horizontal axis wind turbine system mounted on the rooftop of a building at University of California, Davis. The model of this wind turbine is Bergey Windpower XL.1. Compared to the lab tests, the main difference was that the field tests relied on actual wind excitations (as opposed to a controlled motor). In laboratory test, the turbine blade divided into two cases, undamaged and damaged blades. Take two loose bolts connection in one of the turbine blade as damaged cases and use damage index to identify the damage occurrence in blade. In addition, the raw data from both laboratory tests and field tests were analyzed by signal processing and system identification techniques for deriving the model response of the blades. The Multivariate Singular Spectrum Analysis (MSSA), Covariance-driven Stochastic Subspace Identification method (SSI-COV) and Data-driven Subspace Identification method (SI-DATA) were used to identify the dynamic characteristics of the system. These above experiments confirmed that dynamic characterization of rotating wind turbines system was feasible, and the results will guide future wind turbine monitoring studies.口試委員會審定書 # 誌謝 i 中文摘要 ii ABSTRACT iii 目錄 v 表目錄 viii 圖目錄 ix 第一章 導論 1 1.1 研究動機與目的 1 1.2 文獻回顧 2 1.3 研究架構與內容 3 第二章 風機簡述與實驗配置 5 2.1 風機基本振動行為簡述5 2.1.1 風機葉片自然頻率 5 2.1.2 風機葉片轉速頻率 6 2.2 實驗基本配置 7 2.2.1 實驗室試驗之風機簡述及實驗配置 7 2.2.2 現地試驗之風機簡述及實驗配置 8 第三章 結構監測方法 9 3.1 奇異譜分析法 9 3.2 主子空間(Subspace)與零子空間(Null-space)損壞因子識別 11 3.3 隨機子空間識別法 13 3.3.1 狀態空間模型 13 3.3.2 隨機過程 17 3.3.3 隨機子空間識別法流程 18 3.3.4 系統模態特徵萃取 21 3.3.5 系統參數選擇與穩態圖之繪製 22 3.4 子空間系統識別法 23 3.4.1 確定性隨機系統 23 3.4.2 子空間識別法理論 25 3.5 小波分析 27 3.5.1 小波包轉換 28 3.5.2 希爾伯特轉換 29 3.5.3 希爾伯特振幅圖 30 3.6 頻率域分解法 31 第四章 實驗室試驗結果 33 4.1 實驗室試驗(一) 33 4.1.1 實驗案例 33 4.1.2 損壞偵測識別 33 4.1.3 實驗案例分析結果 34 4.1.4 時頻分析法損壞識別結果 38 4.2 實驗室試驗(二)---多角度風機系統 39 4.2.1 損壞偵測識別 39 4.2.2 多角度風機識別結果 40 第五章 現地試驗結果 41 5.1 衝擊力實驗 41 5.2 風機轉動試驗 41 5.2.1 多點奇異譜分析法結果 42 5.2.2 SSI-COV及頻率域分解法識別結果 42 5.2.3 時頻分析法識別結果 44 5.3 實驗結論 46 第六章 結論與未來展望 47 6.1 結論 47 6.2 未來展望 48 參考文獻 50 附表 56 附圖 60 附錄一 114 附錄二 118 附錄三 1235067384 bytesapplication/pdf論文公開時間：2019/07/29論文使用權限：同意無償授權協方差隨機子空間識別法奇異譜分析法風機系統子空間識別法結構健康監測損壞偵測thesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/260867/1/ntu-103-R01521205-1.pdf