Enhanced drive‐by bridge modal identification via dual Kalman filter and singular spectrum analysis
Jiantao Li, Xinqun Zhu, Jian Guo
Abstract
The drive-by bridge health monitoring is to assess the bridge condition using the acceleration responses measured on the body or axle of instrumented vehicles. The vehicle responses are greatly affected by the road surface roughness that makes the bridge dynamic information blurred. Instead of direct using vehicle responses for the bridge monitoring, the dynamic response of contact point (CP) between the vehicle and bridge is further explored to enhance the drive-by bridge modal identification. A novel three-step framework is proposed to extract the components related to the bridge vibration from vehicle responses. The first step is to identify the input forces of two successive vehicles by solving the combined state-input estimation problem using dual Kalman filter. The CP responses of two contact points are calculated using the input forces and vehicle parameters. In the second step, the subtraction technique is applied to the identified CP responses of the two instrumented vehicles and the effect of the road surface roughness can be significantly reduced. Finally, an automatic singular spectrum analysis technique (auto SSA) is incorporated to decompose the response residual. Then the mono-component modes related to the bridge response are extracted from the response residual for drive-by bridge modal identification and/or the nonstationary characteristic identification of vehicle–bridge interaction (VBI) system. Results of numerical and experimental study demonstrate that the method can significantly suppress the vehicle response component and reduce the effect of road surface roughness to enhance the bridge modal identification.