ABSTRACT
ABSTRACT: While we analyze the tool wears signal in turning process, the mode mixing problem caused by Empirical Mode Decomposition (EMD) is a great challenge. In this paper the wavelet analysis combined with the EMD was chosen to extract the features of the cutting tool wears signal. Firstly, the EMD method was used to decompose the wear signal into several Intrinsic Mode Functions (IMF), and wavelet analysis was employed to decompose the IMF and to make it more relevant to the tool wear, the IMF was rebuilt. After calculating the correlation coefficient of the rebuilt IMFs and cutting tool wear, sensitive singles based on the rebuilt IMFs are selected as the inputs of Support Vector Machine (SVM). Finally, we can identify the tool wear condition.
