Prediction of Surface Roughness for HSM Based on BP Neural Network

摘要

A predictive model is presented for the surface roughness in high-speed milling of P1.2738 (plastic die steel) based on BP Neural network. The data for establishing the model is derived from the experiment conducted on a high-speed 5-axis machining center by factorial design of experiments. Compared with measured data and data from regression analysis, the result of prediction using BP neural network indicates its feasibility, which provides reference for the optimization of cutting parameters.