Abstract Vibration-based damage assessment approaches are conventionally and widely adopted to monitor the structural health of a building during the bases excitation process. This study explores the possibility of using structural stiffness parameters to locate the storeys of building that are damaged in an earthquake event. A damaged storeys usually exhibits stiffness parameters variations during a strong enough earthquake, and the corresponding stiffness of storey often change in the earthquake. The wavelet packet transform is applied to the measured acceleration responses of a structure to reconstruct the autoregressive with exogenous input (ARX) model in wavelet packet domain. The modal parameters of the structure are estimated directly through the identified coefficient matrixes of the ARX model. Next, the stiffness parameters of the structure that could be reconstructed form the identified natural frequencies and mode shapes would be used to detect the damage locating of a building. The accuracy of this procedure is numerically confirmed, the effects of the wavelet parameters and noise on the ability to accurately estimate the dynamic properties are also investigated.
[1]
Hidenori Kimura,et al.
Recursive 4SID algorithms using gradient type subspace tracking
,
2002,
Autom..
[2]
Feng Ding,et al.
Recursive least squares parameter identification algorithms for systems with colored noise using the filtering technique and the auxilary model
,
2015,
Digit. Signal Process..
[3]
C. M. Wen,et al.
A neural network approach for structural identification and diagnosis of a building from seismic response data
,
2003
.
[4]
Michael Eichler,et al.
Computing Maximum Likelihood Estimates in Recursive Linear Models with Correlated Errors
,
2009,
J. Mach. Learn. Res..
[5]
Chi-Hung Huang,et al.
STRUCTURAL IDENTIFICATION FROM AMBIENT VIBRATION MEASUREMENT USING THE MULTIVARIATE AR MODEL
,
2001
.
[6]
Lennart Ljung,et al.
Error propagation properties of recursive least-squares adaptation algorithms
,
1985,
Autom..