Search

In this paper, a novel approach of the real-time chatter detection in the milling process is presented based on the scalogram of the continuous wavelet transform (CWT) and the deep convolutional neural network (CNN). The cutting force signals measured from the stable and unstable cutting conditions were converted into two-dimensional images using the CWT. When chatter occurs, the amount of energy at the tooth passing frequency and its harmonics are shifted toward the chatter frequency. Hence, the scalogram images can serve as input to the CNN framework to identify the stable, transitive, and unstable cutting states. The proposed method does not require the subjective feature-generation and feature-selection procedures, and its classification accuracy of 99.67% is higher than the conventional machine learning techniques described in the existing literature. The result demonstrates that the proposed method can effectively detect the occurrence of chatter. [ABSTRACT FROM AUTHOR]

Milling plays a core role in the manufacturing industry. If a milling tool suffers severe wear or breakage during the manufacturing process, it requires immediate attention to prevent precision errors and poor surface quality. Previous studies applied multiple sensors, such as force sensors, microphones, and acoustic emission sensors, to extract the features related to tool wear. However, owing to the complex mechanism causing tool wear, the results of these studies were not only quantitatively different but also varied with respect to cutting conditions. Therefore, this research discussed the relationship between sound signal and tool wear under multiple cutting conditions. Collinearity diagnostics and a stepwise regression procedure were used to optimize the time–frequency statistic features generated from the wavelet packet decomposition. The regression and an artificial neural network model were developed to predict the degree of tool wear. The proposed method provided solid statistical support for the feature selection process, and the results showed that it will maintain prediction accuracy regardless of the different cutting conditions. Further, the method achieved better accuracy than the commonly used root mean square value. [ABSTRACT FROM AUTHOR]