Sensors selection for tool failure detection during machining processes: A simple accurate classification model

Abstract Tool failure detection is a crucial task for continuous safe machining operations. In this study, novel approach proposed to develop an accurate and simple tool condition classification model (TCCM) early during processes. Signals from current, vibration, acoustic emission sensors were preprocessed used feature extraction in both time frequency domains, leading total of 152 features. Next, reduction was carried out based on relative importance, computed using fully-grown random forest, which reduced the number features 15. To find best combination relevant signal features, 32,767 optimized support vector classifiers developed. The comparison between different candidate models accuracy complexity. results showed that up 0.911 attainable process-independent only current sensors. Besides, developing ensemble material-dependent good potential improvement, recording 0.958 while extracted novelty present study its focus single sensor-based high-accuracy TCCM. This opens door wider utilization such technology, especially all existing studies focused increasing multi-sensor TCCMs, increases cost technology makes it inaccessible, small medium enterprises.