Drive-by bridge damage detection based on wavelet analysis of residual contact response of a moving vehicle.

The indirect approaches based on the responses of specially designed moving vehicles on bridges to extract bridge dynamic properties have received great research attention. Wavelet transform (WT) is shown to be a useful tool for the analysis of vehicle responses and bridge damage detection in the vehicle-bridge interaction system. However, past studies have demonstrated that road surface roughness poses difficulty in applying wavelet analysis to successfully detect bridge damage. In this paper, a two-step method is proposed for bridge damage detection based on WT analysis of the residual contact response of the front and rear vehicle wheels of a four degrees of freedom (4-DOF) moving vehicle to reduce the influence of road surface roughness. In the first step, the generalized Kalman filter under unknown input (GKF-UI) developed by the authors is used to identify the unknown contact forces and the state vector of the vehicle. Due to the advantages of GKF-UI, acceleration and displacement sensors can be installed conveniently on the vehicle body instead of on the wheel axles in previous studies. Then, the contact-point bridge deflection responses of the front and rear vehicle wheels are estimated from the identified contact force and vehicle state. In the second step, the residual contact-point bridge deflection response is obtained by subtracting the front contact-point bridge deflection and rear contact-point bridge deflection with a time shift to reduce the influence of road surface roughness. Then, the locations of bridge damage can be identified based on WT of the singularities of residual contact-point bridge deflection response. Numerical examples of bridge damage detection demonstrate the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]