Uncertainties qualitative and quantitative analysis of machining errors of thin-walled parts under kriging framework

Abstract

The machining of aviation thin-walled parts has always been a difficult task due to their high machining accuracy requirements. During the machining process of thin-walled parts, machining errors are inevitable, reducing the precision and quality of the components. Therefore, predicting machining errors is beneficial for optimizing process parameters, improving accuracy, and reducing costs. This paper proposes a thin-walled parts machining errors prediction method based on the kriging regression model considering uncertainties. The method establishes a surrogate model considering process parameter uncertainty, part uncertainty, model uncertainty, and unquantified uncertainty. The experimental data were extracted through processing experiments, combined with a Bayesian estimation algorithm to ascertain the model parameter, and combined with a gradient descent method to iteratively solve the uncertainty in the model to determine the optimal parameters of the model. The result of the experiment shows that the prediction model has a higher accuracy in the training parameter range and also in the extrapolation test set, and the root mean square error (RMSE) of the prediction model is 5.4238 μm.