Bayesian Inference of Mechanical Property Parameters and Layup Orientation Angle in a Composite Wing

A Bayesian method is adopted to infer the mechanical property parameters and layup orientation angle of a composite wing in this paper. In order to make the calculation feasible, a basis-adaptive PC-Kriging model is proposed, which can replace the finite element model of the composite wing accurately. Considering that the simulation model of the composite wing is a high-dimensional model, AIES algorithm is used to model calibration of the composite wing, because it shows good performance in processing high-dimensional strongly correlated posterior distribution parameters. Finally, comparing the predicted parameters with the real parameters, the error is found to be small, which indicates that the proposed Bayesian method has good performance.