A PCA-Integrated OGM (1, N) Predictive Model for In-Process Tool Wear Prediction Based on Continuous Monitoring of Multi-Sensorial Information

In the modern machining process, catastrophic tool failure has become a significant problem because the machining quality and efficiency depend on the condition of the tool. To know the real-time tool conditions, a tool wear monitoring (TWM) system can be developed based on the continuous monitoring of in-process multi-sensor signals without interrupting the machining process. In this investigation, a novel PCA-integrated optimization grey prediction OGM (1, N) model has been proposed to develop a TWM system. During the machining process, machined surface texture and cutting tool vibration data have been acquired. Gabor wavelet transform (GWT) has been used for the extraction of the features, and the PCA technique was used for dimensionality reduction and most significant feature selection. Finally, a novel PCA-integrated OGM (1, N) predictive model was trained with the selected features and predicted real-time tool wear with 94.85% of prediction accuracy.