Cutting tool wear identification based on wavelet package and SVM

Abstract

By contrast with conventional methods, Acoustic Emission (AE) sensor possesses better performance for tool wear identifying. So, AE sensor is employed into cutting tool wear identification in this paper. Because of the diversity and time varying of AE, wavelet package decomposition and Support Vector Machine (SVM) are employed to process AE signal. Wavelet package is suitable for analyzing non-stationary signal, and SVM possesses excellent classification capacity for small sample. According to these features, signal processing method for identifying fault of cutting tool wear based on wavelet package and SVM was presented. The characteristics of the cutting tool wear under different conditions were extracted by wavelet package, and cutting tool wear was identified by SVM classifier. Experiment results show that the method based on wavelet package and SVM is suitable for identifying cutting tool wear, and the rate of successfully identifying is 93.3%.

Tool Wear Identification in Turning Titanium Alloy Based on SVM

“…But as literatures [5,6] show: SVM has superior toolwear monitoring performance to HMM and neural network in case of small sample. With the advantages of strong nonlinear mapping ability, small samples et al , SVM is widely used in tool state recognition , Bulent Kaya [7] used SVM trained with the features that were selected with genetic algorithm (GA) to monitor tool states and got 100% tool state recognition rate , Xu tao et al [8,9] got above 90% accuracy using SVM .…”