Empirical fragility quantification model of regional building portfolios considering optimized seismic intensity measures

Abstract

Seismic fragility and hazard analyses of large-scale zonal building portfolios are fundamental for evaluating the earthquake performance of rural and urban structure clusters. Using probabilistic hazard and risk theory, an updated probability and risk model is developed to effectively display the seismic fragility features of regional buildings. However, the conventional definitions of vulnerability level and macroseismic intensity have features of fuzziness and randomness, resulting in relatively low quantitative accuracy of the fragility of structures in large-scale regions. This paper improves upon the traditional quantitative indicators of vulnerability level and develops an innovative scale to assess the four typical structure clusters in large-scale regions. The conventional macroseismic intensity measurements and logarithmic fitting model are updated using earthquake hazards, probability and reliability, the Bayesian model, and risk theory. The ground motion parameters and instrument seismic intensity measurements are auxiliary quantitative indicators. An improved vulnerability model of large-scale zonal structures considering updated intensity measures and an optimized nonlinear regression method is proposed. The correctness of the developed model is verified via actual seismic loss survey data (18,622.6086 ×10⁴ m²) from 57 powerful earthquakes that occurred in Xinjiang, China, from 1993 to 2023. By combining the vulnerability index model and data-cloud algorithm, a fragility curve and quantitative strip model for updating fragility databases considering large-scale zonal structure clusters are developed.