Prediction of Tool Remaining Useful Life of NC Machine Tool Based on DTW Algorithm and LSTM Neural Network

Based on dynamic time warping (DTW) algorithm and long short-term memory (LSTM) neural network, this paper carried out research on the prediction of tool remaining useful life (RUL) of NC machine tools, evaluated the accuracy of the health index prediction model of each tool by using root mean square error, and introduced a scoring mechanism to score the predicted results of tool residual service life based on the specified prediction starting point. The method can realize the prediction of the tool remaining useful life.