Monitoring tool wear states in turning based on wavelet analysis

To monitor the tool wear states in turning, a new way based on the wavelet transform was proposed to get the signal characters, which can reflect the tool wear states. Using discrete dyadic wavelet transform, the acoustic emission (AE) signal of cutting process was decomposed; the root mean square (RMS) values of the decomposed signals at different scales were taken as the feature vector; the technique of fuzzy pattern identification was used to identify the tool wear states. Experimental results showed that the monitoring technique based on wavelet analysis is suitable for real-time implementation in manufacturing application.