Design and Analysis of RUL Prediction Algorithm Based on CABLSTM for CNC Machine Tools

In the cutting manufacturing process, cutting tools are one of the factors closely related to the efficient production and quality of products. order to extend the life of the cutting tool, it is not possible to efficiently set the cutting conditions, resulting in reduced product productivity. Conversely, it is very important to predict the remaining wear time of the cutting tool. If the cutting conditions are set to increase product productivity, the cutting tool wears quickly, and if damage cannot be sustained, the corresponding line stops, so it is very important to predict the remaining wear time of the cutting tool. However, it is difficult to predict the tool's RUL (Remaining Useful Life) because various are linked together due to the cutting conditions. In this paper, algorithms such as CNN, BILSTM, etc. are used to predict the RUL of a CNC machine tool using the BiLSTM-based CABLSTM algorithm and advanced Taylor's tool life equations for factors affecting complex RUL.