Vibration signature based stability studies in internal turning with a wavelet denoising preprocessor

Abstract

Detection of accurate chatter is one of the significant problems in machining science due to the contaminated noise present in the acquired signals, which reduces the efficiency of the signal processing tool. In the present work, an efficient chatter detection approach based on the combined modified wavelet de-noising (WD) and Hilbert-Huang transforms (HHT) is adopted. The proposed methodology is implemented in four stages: (i) de-noising of contaminated signals using novel adaptive threshold-based wavelet de-noising, (ii) analysis of de-noised signals with Hilbert-Huang transform to obtain the time-frequency spectrum, (iii) extraction of features such as mean, standard deviation, root mean square and kurtosis of the sensitive intrinsic mode function (IMF) to identify the chatter onset and (iv) classification of chatter onset states based on statistical measures using an improved probabilistic neural network (PNN). The proposed approach is implemented for identifying the cutting states of experimentally measured vibration signals from internal turning operation.