Clustering‐based adaptive ground motion selection algorithm for efficient estimation of structural fragilities

To accurately predict the seismic demands of structural systems, a proper set ground motions representing hazard given site is needed. In general, such includes large number motions, and thus may result in high computational cost. address this challenge without compromising accuracy fragility, paper proposes clustering-based algorithm that can select representative subset adaptively from motions. First, critical features significantly affect system are identified by Lasso regression responses various single degree freedom systems on existing intensity measures Second, selected based hierarchical clustering until fragility curve converges. Applications to reinforced concrete building steel moment-resisting frames demonstrate improved efficiency wide applicability proposed method. The results numerical examples confirm robust performance against types, definitions limit-states. enables us obtain fragilities using reduced while keeping consistency with available motion set.