Probabilistic seismic hazard model and vulnerability analysis for typical regional structures.

The estimation of the seismic vulnerability of regional structures has contributed positively to the development of seismic risk and resilience in large-scale building clusters. However, the empirical structural seismic damage and ground motion data used to validate the risk model are relatively outdated, resulting in insufficient evaluation accuracy of the established risk model and an urgent use for retrofitting. This study analysed the actual failure peculiarities of typical structures during a recent typical destructive earthquake in China (the Luding earthquake in Sichuan Province on September 5, 2022). A total of 300,000 acceleration records from ten actual monitoring seismic stations were processed and analysed. Nonlinear dynamic time history curves, response spectra, and combined seismic acceleration cloud images were generated considering different directions. Moreover, the traditional instrument intensity calculation equation has been updated, and an instrument intensity curve bunch considering 300,000 earthquake acceleration records has been developed. To study the probabilistic seismic risk of large-scale regional buildings, the developed vulnerability database of typical regional structures (Sichuan Province) was supplemented and optimized, and an updated vulnerability database (413 damage probability matrix) was established, considering the failure of five structures in 32 destructive earthquakes. An updated structural seismic vulnerability model based on extended vulnerability grades was proposed, and a mean vulnerability probability matrix based on structural risk datasets was established. A membership index function for estimating the correlation between different vulnerability grades was innovatively proposed, and the rationality of the model was verified using an improved empirical dataset. A multidimensional failure prediction model was developed based on an improved mean seismic damage index. [ABSTRACT FROM AUTHOR]