Application of Singular Value Decomposition Technique to Strong Ground Motion Data Analysis
Date Issued
2007
Date
2007
Author(s)
Tzeng, Sheng-Hao
DOI
en-US
Abstract
The theme of this paper is to provide a relatively simple and straightforward procedure to compute the probabilistic dynamic response of linear SDOF system using non-stationary excitation data by applying singular value decomposition method (SVD).
After compute the probabilistic dynamic response, we can regard it as a reference response to the specified area (such as Taipei basin or Chia-nan plain) and the specified earthquake event (such as 1999_9_21 or 2002_3_31 earthquake event). Then we can analysis the difference between reference and records at each station. The result can be further compared with current seismic code. And we can make some comments and suggestions to it.
In a traditional approach, the statistical analysis is commonly used. The response of each record is derived first. After that, the response spectrum corresponding to each earthquake records were computed. Then, the mean and standard deviation of these response spectra corresponding to each station and each event may be derived. The statistical analysis requires a huge amount of data and computation takes time. The proposed method will alleviate this aspect of problems.
Followed by the procedure described in the thesis, three analysis case were used in this study, (i) 2002_3_31 in Taipei basin. (ii) 1999_9_21 in Taipei basin. (iii) 1999_9_21 in Chia-nan plain. According to the analysis result and current seismic, the reference time history is quite enough to represent the behavior of the specified area and earthquake event. Besides, a revision to some parts of seismic zone in Taipei basin might be needed.
Subjects
奇異值分解法
singular value decomposition
Type
thesis
File(s)![Thumbnail Image]()
Loading...
Name
ntu-96-R92521243-1.pdf
Size
23.31 KB
Format
Adobe PDF
Checksum
(MD5):59dfb310486529faac7892bee17972b9