Please use this identifier to cite or link to this item:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/28517| Title: | 利用基因演算法標定地震波相相對到時的探討 Relative seismic travel time determined by the genetic algorithm and the micro genetic algorithm |
| Authors: | Shang-Hsueh Yu 余尚學 |
| Advisor: | 喬凌雲(Ling-Yun Chiao) |
| Keyword: | 基因演算法,相對到時, genetic algorithm,micro-GA,relative arrival time, |
| Publication Year : | 2007 |
| Degree: | 碩士 |
| Abstract: | 地震波抵達區域地震網後各站的相對到時是決定地下速度構造的重要依據,如何取得正確的相對時間是非常重要的工作。
以往處理地震相對到時的問題時,是將地震相對到時認為是超定的(overdetermined)線性情況下求出一組相對到時最近似解;而本研究則是將基因演算法(genetic algorithm)具處理非線性的問題的能力,應用於地震相對到時問題。 本研究以IndepthIII與海底地震儀資料進行實驗。在實驗的同時,也同時將傳統基因演算法與微型基因演算法在相同的參數下進行實驗,來比較何者效能較優。 在海底地震儀的實際應用實驗中,主要以P波與S波兩種波相作為相對時間的測試對象;資料處理時,將地震資料事先經過歸一化(normalization)、濾波(filter)與排列(alignment)後,對於基因演算法運算的速度與實驗結果會有明顯正面的幫助。 從實驗的結果發現,基本上只要能夠定義出波相的範圍與位置,基因演算法即能標定地震網內各站的正確相對時間。 Abstract Seismic relative arrival times among a local network provide important information for the regional velocity structure. To map the precise velocity structure, we need to find precise relative arrival times. In the past, determining seismic relative arrival times is posed as an overdetermined and linear inverse problem to find an approximate solution. In our method, it is formulated in terms of a non-linear problem and is dealt with using the genetic algorithms. We experiment with data from IndepthIII and an Ocean Bottom Seismometer array. We compare the efficiency between the genetic algorithm and the micro-genetic algorithm. It is shown that after proper normalization、filter and alignment. GA is very efficient in determining the relative arrival times. |
| URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/28517 |
| Fulltext Rights: | 有償授權 |
| Appears in Collections: | 海洋研究所 |
Files in This Item:
| File | Size | Format | |
|---|---|---|---|
| ntu-96-1.pdf Restricted Access | 1.34 MB | Adobe PDF |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.
