Bayesian Model-Updating Implementation in a Five-Story Building

Simplifications and theoretical assumptions are usually incorporated into the numerical modeling of structures. However, these may reduce accuracy simulation results. This problem has led to development model-updating techniques minimize error between experimental response modeled structure by updating its parameters based on observed data. Structural models typically constructed using a deterministic approach, whereby single best-estimated value each structural parameter is obtained. often complex involve many uncertain variables, where unique solution that captures all variability not possible. Updating Bayesian Inference (BI) have been developed quantify parametric uncertainty in analytical models. paper presents implementation BI five-story building model quantification associated uncertainty. The framework implemented update calculate covariance matrix output information provided modal frequencies mode shapes. main advantage this approach data considered defining likelihood function as multivariate normal distribution, leading better representation actual behavior. results showed effectively allows statistically rigorous parameters, characterizing increasing confidence model’s predictions, which particularly useful engineering applications critical.