Material removal profile and surface topography prediction of ultrasonic vibration-assisted polished based on gray wolf optimization neural network

Ultrasonic vibration-assisted polishing (UVAP) has important research value in improving polishing efficiency and polishing quality. The material removal function also has an important contribution to the optimization of process parameters and prediction of surface topography. In previous studies, few studies combined UVAP material removal functions with intelligent algorithms. Therefore, a prediction method of the UVAP material removal model (MRM) based on the GWO-BP algorithm is proposed in this paper. The material removal profile (MRP) was predicted using the GWO-BP algorithm based on the established UVAP material removal model for convex and concave cylindrical polishing tool tilts. The results show that the material removal model predicts the material removal depth (MRD) and material removal rate (MRR) with an error of less than 10%. Compared with the traditional BP algorithm, the average error of the GWO-BP algorithm prediction was reduced by more than 20%, and the prediction error of the surface Sa value was 5.51%. This method provides a good guide for analyzing the mechanism of UVAP and the rapid prediction of machined surface topography.