Bayesian Updating Framework for 3d Fem Building Models Using TMCMC

This paper proposes a Bayesian updating framework for three-dimensional finite element building models based on Transitional Markov chain Monte Carlo (TMCMC). The method employs a multi-step scheme to robustly update a large number of model parameters. The framework is applied to shaking-table tests of a six-story reinforced-concrete residential building specimen, and its performance is examined through comparison with the measured responses. The results show that the updated model reproduces the observed accelerations in all three directions with good accuracy, demonstrating the practicality of the proposed TMCMC-based Bayesian updating approach for complex 3D building models.