Rapid Seismic Damage Assessment of RC Bridges Considering Time-Frequency Characteristics of Ground Motions

The traditional seismic damage assessment involved a mass of filed investigations conducted by engineers or researchers, genetically to be time-consuming and inefficient. With the wide application of deep learning technique to earthquake engineering, it has been found out to be a powerful tool having great potential in damage identification after disaster events. This study proposed a hybrid method to evaluate seismic damage of bridges considering time-frequency characteristics of ground motions, in which the ground motion features were explored by wavelet transform technique, and optimized CNN models were utilized to establish the mapping relationship between structural damage and characteristics of ground motions. Compared with the existing damage assessment method, this proposed framework emphasized the fact that structural responses were time-frequency-dependent; moreover, damage states were quantitatively identified by more reasonable demand parameter, providing not only the degraded status but also the structural ductility. This method was demonstrated and verified by a typical RC continuous bridge as a case study, and comparing with the existing method, the recognition precision was significantly improved from 76.2 to 83.3% and recall from 83.3 to 95.2%, indicating the positive effects of performance-based criteria on seismic damage assessment.