Vibration-based Structural Dynamic Identification and Damage Detection
Date Issued
2016
Date
2016
Author(s)
Lee, Chung-Hsien
Abstract
Development of structural health monitoring in civil engineering, including system identification and damage detection on structure system, is popped up in recent decades. Due to the emerging sensing technology many structural monitoring data can be obtained. In this thesis researches will focus on using out-put only data analysis to extract features from structural responses. Data collected in this study will include measurement from acceleration, angular velocity and strain data. Based on the stochastic subspace identification algorithm (SSI), multivariate AR model the dynamic characteristics of the study structures can be identified. Damage detection algorithms, including migration of AR coefficients, ....., are also developed for damage localization. Besides, in this study data collected from Gyroscope is also used for system identification and damage detection. Methodology on the estimation of floor permanent deformation using angular velocity and blind source separation technique is also developed. Verification of the proposed methods by using the shaking table test of two 3-story steel structures subjected to a series of white noise excitation back to back after each earthquake excitation
Subjects
structural health monitoring (SHM)
covariance-driven stochastic subspace identification (SSI-COV)
multivariate autoregressive model (MVAR)
continuous wavelet transform (CWT)
singular spectrum analysis (SSA)
angular velocity analysis
microelectromechanical systems (MEMS) gyroscopes
Type
thesis
File(s)![Thumbnail Image]()
Loading...
Name
ntu-105-R03521230-1.pdf
Size
23.32 KB
Format
Adobe PDF
Checksum
(MD5):9e5e50a6502a07355ab45e6859dc738d