Damage detection for bridges under a moving vehicle based on generalized S - local maximum reassignment transform

Abstract

The dynamic responses of bridges under operational traffic loads exhibit time-varying characteristics, and accurately identifying the time-varying characteristics is crucial for the health monitoring of bridges. In this study, a generalized S-local maximum reassignment transform (GS-LMRT) is proposed for identifying the time-varying frequencies of bridges under a moving vehicle and for detecting the damage in bridges. First, the parameter optimization algorithm using the energy concentration measure is adopted to determine the optimal window parameters for the generalized S transform (GST). Then, by introducing the frequency-reassignment operator from the local maximum synchrosqueezing transform (LMSST) to the absolute value of GST, the GS-LMRT can be obtained. GS-LMRT can improve the energy concentration of the GST and overcome the issue of fixed window width in LMSST. Finally, to assess the damage of bridges, a damage index (DI) is calculated using the IFs obtained from intact bridges and damaged bridges under a moving vehicle through the proposed GS-LMRT method. Numerical investigations are conducted to study the effect of vehicle mass, vehicle speed, measurement noise, and road surface roughness. Laboratory experiments of a damaged beam with a moving vehicle are designed to validate the proposed method. Numerical simulation and experimental results illustrate that the proposed GS-LMRT method can be used for instantaneous frequency identification and damage detection of bridges under a moving vehicle.