Finite element model updating of building structures under seismic excitation: A parallelized latent space-based Bayesian framework

Abstract

Enhancing seismic fragility and risk assessment of nuclear power plants relies on accurate prediction of reactor building responses to seismic hazards, which can be further improved through dynamic analysis of high-fidelity finite element (FE) models. However, FE models often exhibit non-negligible discrepancies from actual structures due to various sources of uncertainty, necessitating FE model updating with rigorous quantification of associated uncertainties. This paper presents a GPU-accelerated latent space-based Bayesian framework for FE model updating of building structures. In the proposed framework, high-dimensional structural response data (e.g., time histories or frequency response functions) are projected into a low-dimensional latent space using a multimodal variational autoencoder (MVAE), thereby enabling efficient and tractable likelihood evaluation without explicit modeling in the original observation space. Once trained, the surrogate enables amortized inference, allowing posterior sampling to be performed without additional simulator evaluations. We specifically employ a sequential Monte Carlo (SMC) sampler, whose population-based formulation allows parallel evaluation of the approximate likelihood on GPUs, resulting in computational efficiency and robustness against multimodal and complex posterior distributions. The proposed framework is evaluated through numerical benchmarking using a shear-type building structure with a multimodal posterior distribution. The results indicate that the proposed approach can recover the major posterior modes while achieving efficient inference through GPU-based parallel evaluation. We further apply the framework to experimental data from a shaking table test of a reinforced concrete building structure. The results suggest that the inferred posterior distributions and posterior predictive responses are consistent with the observed data, even in the presence of sparse observations that induce complex posterior distributions.