A multi-level data-driven Bayesian approach for stochastic model updating of complex aeroelastic systems

McGurk, Michael and Yuan, Jie (2024) A multi-level data-driven Bayesian approach for stochastic model updating of complex aeroelastic systems. In AIAA SCITECH 2024 Forum. Aerospace Research Central. 13 pp . (doi:10.2514/6.2024-0193).

Record type: Conference or Workshop Item (Paper)

Abstract

The objective of this work is to propose a multilevel data-driven Bayesian framework to minimise the sampling requirement for stochastic model identification of general nonlinear aeroelastic systems. Adaptive Kriging-based surrogate models are developed through multi-level Bayesian modelling updating to represent the limit cycle oscillation (LCO) response of the nonlinear aeroelastic model. The proposed methodology is demonstrated on an aerofoil wing flutter model together with LCO experimental testing data. The results is benchmarked by the counterpart from a classical single-level approach. It is observed that the proposed approach offered a 74% reduction in training data requirement/run time and 3% accuracy improvement for the surrogate model.

Text

Scitech_final_paper (7) - Accepted Manuscript

Available under License University of Southampton Accepted Manuscript Licence.

Download (5MB)