Editorial: Recent advances in stochastic model updating

As a classical technology, Model Updating has been developed for nearly 50 years to calibrate the parameters or the numerical model itself, such as to tune its prediction as close as possible to experimental measurements. Industries have benefitted from a more precise model, which further promotes the application of numerical simulation technologies, such as the finite element method and computational fluid dynamics. However, it is widely recognized that the unavoidable uncertainties in both operational experiments and numerical analyses must be understood by the process of model updating. Uncertainty analysis has enabled the progression of model updating from the deterministic domain to the stochastic domain. Non-deterministic modelling approaches enable characterization, propagation, and quantification of the inevitable uncertainties, providing predictions over a possible range of outcomes (distributional, interval, fuzzy, etc.) rather than a unique solution with maximum fidelity to a single experiment. Such approaches provide confidence in structural dynamics, and computer-aided engineering generally, backed up by detailed uncertainty quantification.