Prediction of milling force based on multiple linear regression and BP neural network

In order to predict the milling force, getting the fitting curve of the predicted values and experimental values, linear regression milling force prediction model and BP neural network were established. And then through the linear regression theory to remove the abnormal milling force data points that acquired by experiment, and train the BP neural network prediction model to guide into experiment datum and then predict milling force, at last, the predictive value and curve fitting of experimental value were acquired. The results show that the BP neural network forecasting model is more suitable to milling force, and through the theory of linear regression to remove abnormal points, which make the BP neural network prediction more accurate. © 2014 Binary Information Press.