This paper presents a novel framework for Bayesian structural model updating and proposes a method that utilizes the surrogate unimodal encoders of a multimodal variational autoencoder. This method facilitates an efficient nonparametric estimation of the likelihood describing the observed data. It is particularly suitable for high-dimensional correlated simultaneous observations applicable to various dynamic analysis models. The proposed approach is benchmarked using a numerical model of a single-story frame building with acceleration and dynamic strain measurements.
翻译:暂无翻译