以HHT 法進行民用飛機飛行資料分析之研究
學年 101
學期 1
出版（發表）日期 2012-08-01
作品名稱 以HHT 法進行民用飛機飛行資料分析之研究
作品名稱（其他語言） Airline Quick Access Recorder (Qar) Data Analysis via the Hht Method
著者 宛同
單位 淡江大學航空太空工程學系
描述 計畫編號：NSC101-2221-E032-007
 研究期間：201208~201307
 研究經費：641,000
委託單位 行政院國家科學委員會
摘要 在本研究計畫中，現代的資料分析工具希爾伯特黃轉換(Hilbert-Huang Transform, HHT) 被用於分析飛行品質保證系統飛行資料之時、頻和能量頻譜。我們先前的研究中，HHT 被 證實可以用於2-D 拍撲翼的計算流力結果之檢驗分析。HHT 主要包含兩個部分：經驗模態 分解(EMD)及希爾伯特頻譜分析(HSA)。EMD 是一個過濾工具，其能分解非線性及非定 性資料，這些被分解後的子訊號稱之為本質模態函數(IMF)，可以提供我們一個更深入的 物理或統計上的視野。儘管不是每一個IMF 都富含物理意義，但含有數學上的意義，可以 用於不同資料列的比較，FOQA 的飛行資料亦是如此。每把航班的相同參數的IMF 特性遵 循著類似的特性，故本計畫試著探討如何找到該特性，用於事先避免飛行意外的產生。我 們建立了一個叫做基因譜的工具，將一個所有的統計資料(含IMF)紀錄在一個DNA 序列， 可以用於檢驗和比較。傳統的資料分析無法比較數列的本質層面，是故使用基因譜來輔助 分析。我們期望HHT 可以賦予我們一探飛行本質的機會，分析的結果用於改進整套FOQA 的系統。研究結果將有助於我國飛安水準的繼續提昇。 Objective in this project is to investigate the hidden time-energy-frequency behavior of Flight Operational Quality Assurance (FOQA) data sets by the aid of a modern, superior data-analysis tool called Hilbert-Huang Transform (HHT). In our previous study, we found out HHT has the ability to examine 2-D flapping wings CFD results. HHT consists two fundamental parts: the Empirical Mode Decomposition (EMD) method and the Hilbert Spectral Analysis (HSA). EMD is a sifting process that has the ability to decompose any non-linear and non-stationary signals to finite components called Intrinsic Mode Functions (IMF) which we are interested in, providing us a more physical insight of the data as well as statistical significance. A FOQA parameter data set usually contains several components that may be distinguish and may have physically meaningful information in it. Despite IMF data set may not be physically meaningful it is still valuable for the use of identification since there exists a similar IMF pattern of each parameter per flight. We establish a new principle called Genetic Profile for HHT identification so that the external and internal appearance can be classified into many statistical subsequences. The motivation of generating genetic profile for HHT results is majorly because the lack of overall perspective on IMFs and marginal spectrums. The specified DNA (1 DNA, 1 Flight) sets of a sequence of flights can tell us how the parameter changed in time. It is hard to tell the whole story by only investigating parameters literally instead of seeing the completeness behind the data. HHT gave Airlines/FOQA engineers a great chance to take a glance on flight essence and discovered great physical meanings among IMFs to improve the entire FOQA system. It is expected that our results will be quite valuable to airlines safety monitoring office and will be able to further enhance the aviation safety standard.
關鍵字 HHT 法; 飛行資料; 異常訊息; Hilbert-Huang Transform Method; Flight Data; Abnormal Signal
語言 zh_TW
相關連結

機構典藏連結 ( http://tkuir.lib.tku.edu.tw:8080/dspace/handle/987654321/103071 )

機構典藏連結