Prediction of maximum deformation of single nail riveting based on BP neural network

Abstract

Riveting deformation cannot be accurately predicted during aircraft assembly, and thus the BP neural network was applied to the field of prediction of the maximum riveting deformation. First, Python language was used for secondary development of ABAQUS to carry out parametric modeling and batch processing. Secondly, a sample size of 500 was designed with the Latin hypercube sampling (LHS) method, of which 450 samples were used for training and 50 for testing. Finally, with the rivet diameter, the length of the shank, the aperture of the plate and the displacement of the pressure riveting as input variables, and the maximum deformation as the output, a BP neural network prediction model was established to predict the maximum deformation of a single screw riveting. The results show that the average prediction accuracy of the model reaches 95.86% in the prediction of the maximum deformation of a single nail riveting.

Keywords

• single nail riveting
• secondary development of abaqus
• parametric modeling
• bp neural network
• maximum deformation prediction