Machine Learning Approach to Tool Wear Prediction Using Spindle Motor Signals in Milling Machine

Tool wear monitoring is an essential element in ensuring product quality, preventing damage to the machine, and optimizing the effectiveness of production during manufacturing processes. This paper proposes a machine learning-based model to monitor tool wear using spindle motor AC and DC signals collected during the machining process. This study uses NASA milling machine data from different machining operations under various conditions. Features are extracted from the raw signal, encompassing both time-domain and frequency-domain characteristics. These extracted features are applied to train and validate diverse machine learning models, including approaches like linear regression (LR), random forest (RF), and support vector regression (SVR). The study uses mean error and root mean square error to compare the performance of different models to find the most accurate tool wear prediction algorithm. This will help with monitoring tool wear in milling processes in real-time