Analysis of traffic-induced vibration and damage detection by blind source separation with application to bridge monitoring
Sheng‐Fu Chen, Tzu-Yun Hung, Chin‐Hsiung Loh · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2015
The objective of this study is to demonstrate the application of two different system identification methods on the structural health monitoring of a bridge. The numerical simulation of bridge-vehicle interaction with road surface roughness is considered in this study for system identification. To identify the bridge dynamic characteristics Covariance-driven Stochastic Subspace Identification method (SSI-COV) in cooperated with Wavelet Packet Transform (WPT) decomposition are used to extract the natural frequencies and mode shapes of the system. For comparison, a popular blind source separation technique called Second Order Blind Identification (SOBI) is also used. Comparison between these two different identification methods is discussed. It was demonstrated that the bridge natural frequencies can be identified by the proposed two system identification techniques. Besides, the SOBI algorithm can avoid the difficulty of determining of parameters by using SSI-COV algorithm, such as system order, row of Hankel matrix, etc. Finally, a damage scenario of the bridge structure is provided and damage detection algorithms are also proposed to quantify and locate the damage.