Resolving uncertainty in mechanical vibration of thin structures: a novel Bayesian approach

Over the recent decades, the powerful framework of Bayesian inference has found applications in many areas of science and engineering, rooting itself as an invaluable tool to solve problems where classical deterministic methods fall short. Utilising measured data in conjunction with prior knowledge about the system in question in this Bayesian setting has birthed many robust techniques capable of performing uncertainty quantification, parameter estimation, model selection and more. In this talk, a novel adaptation of the conventional approximate Bayesian computation-sequential Monte Carlo (ABC-SMC) algorithm will be presented. Methodology of this type is used as a means for parameter estimation when there does not exist a tractable form of the associated likelihood function, which is replaced by a simulating model capable of producing artificial data. The aforementioned methodology has been utilised by the authors in the context of resolving uncertainty in measured data obtained from the mechanical vibration of thin structures, and demonstrates the value of this methodology when considering inference problems of this type. By attending this talk, participants will enhance their understanding of how Bayesian methods provide sound mechanisms for parametric inference in an engineering context, encouraging the use of this compelling machinery in their own research.