Mode Shape Identification Using Drive-by Monitoring: A Comparative Study

The structural system identification technique referred to as drive-by health monitoring has recently become popular for the dynamic investigation of bridges. Drive-by monitoring enables extracting dynamic characteristics of bridges such as modal frequencies, damping, and mode shapes by using vibration responses measured on sensors mounted on a passing vehicle. Due to the mobility of the vehicle, the sensor receives the dynamic information of each point along the bridge and allows for acquiring higher resolution in the mode shapes compared to the conventional direct monitoring where a limited number of sensors can be mounted on a bridge. Therefore, theoretically, drive-by monitoring systems promise better resolution in constructing mode shapes, which will, in turn, yield better accuracy in many engineering applications such as damage detection, model updating, and calibration. To this end, this chapter aims at investigating two of the most used drive-by identification methodologies to obtain mode shapes of bridges. In this numerical study, the mode shapes of a single-span bridge are computed from the dynamic response of the instrumented vehicle by applying (i) variational mode decomposition together with the amplitude histories of the Hilbert transform and (ii) a system identification algorithm, namely, the reference-based Stochastic Subspace Identification. The efficacy of the two methods in mode shape identification is investigated using a parametric study, where the sensitivity of the results to different vehicle speeds, sampling rates, and road surface roughness is evaluated.