Gaussian process regression-based surrogate modelling for direct loss-based seismic design of low-rise base-isolated structures

Seismic base isolation has gained popularity in the last decades. As a result, many structures are now equipped with systems to offer enhanced seismic performance and meet needs of risk-aware stakeholders. However, robust performance-based design these types is generally not carried out due iterative nature common approaches time/computational resources required for such iterations, which incompatible preliminary phase. Indeed, risk/loss often just assessed at end process as final verification step. This paper offers an overview simplified methodology low-rise system achieve predefined level earthquake-induced economic loss while complying minimum structural reliability. The main advantage proposed that it requires no iterations. procedure enabled by Gaussian-process-regression-based surrogate probabilistic demand modelling equivalent single degree freedom (i.e., probability distribution peak horizontal displacements accelerations on top layer conditional different ground-motion intensity levels). Combined models non-structural components superstructure, this approach allows mapping range configurations their resulting loss. A designer can then select one identified combinations strength superstructure properties conforming target, reliability requirements, consequently detail accordingly. introduces implemented provides tentative Direct Loss-based Design base-isolated structures.