Prediction of milling force based on spindle current signal by neural networks

Abstract

Milling force is an important physical indicator, which affects chatter stability, tool wear and life. Based on the theory that there is a non-linear mapping relationship between the spindle current and the milling force signal, this paper proposes a method for predicting the instantaneous milling force based on a neural network model of the current signal. The current signal is processed by a sliding window to establish the input signal for the model, and a multidimensional current signal is used to predict the one-dimensional milling force signal. The proposed neural network model is capable of the time lag relationship between the current signal and the instantaneous milling force and reconstructing the instantaneous milling force by combining multiple feature information according to the variation of the current. The experimental results show that the proposed method can achieve an accurate prediction of the instantaneous milling force under different milling parameters.