A wavelet-based characteristic vector construction method for machining condition monitoring

Thin-walled parts are easy to deform during the machining process. The machining conditions of such parts are difficult to monitor since both tool wear and work-piece deformation need to be carefully monitored. Usually, the characteristic vector constructed based on monitoring signals can be deemed as the basis to recognize different machining conditions. In order to construct the characteristic vector, multiple sensors including a dynamometer sensor and an acceleration sensor are used to collect cutting force signals and vibration signals respectively, and the wavelet decomposition method is utilized as the signal processing method for the extraction of signal characteristics including mean and variance of a certain degree of the decomposed signals. Machining experiments are performed to demonstrate and validate the proposed approach.