Surrogate probabilistic seismic demand modelling of inelastic single‐degree‐of‐freedom systems for efficient earthquake risk applications

This paper proposes surrogate models (or metamodels) mapping the parameters controlling dynamic behaviour of inelastic single-degree-of-freedom (SDoF) systems (i.e., force-displacement capacity curve, hysteretic behaviour) and their probabilistic seismic demand model (PSDM, i.e., conditional distribution an engineering parameter [EDP] given a ground-motion intensity measure [IM]). These metamodels allow rapid derivation fragility curves SDoF representation structures. Gaussian Process (GP) regression is selected as metamodelling approach because flexibility in implementation, resulting accuracy computational efficiency. The metamodel training dataset includes 10,000 analysed via cloud-based non-linear time-history analysis (NLTHA) using recorded ground motions. proposed GP regressions are tested predicting PSDM both database (through ten-fold cross validation) eight realistic reinforced concrete (RC) frames, benchmarking results against NLTHA. An application conducted to propagate such modelling uncertainty vulnerability/loss estimations. Error levels deemed satisfactory for practical applications, especially considering low required effort time. Regarding single-building applications enabled by metamodel, this presents first attempt at direct loss-based design procedure, which allows setting target loss level designed structure (shown RC frame). earthquake risk involving exposure vulnerability modules illustrated example building portfolio applications. Specifically, considers retrofit-based risk-reduction policy synthetic portfolio, it possible estimating evolution over