Prognostic Study of CNC Machine Component Using a Systematic Method

The prognostic study of CNC machine component plays a key role in increasing the reliability and safety of manufacturing systems. A systematic prognostic method based on multi-sensor information for CNC machine component is proposed in this paper, where the data characteristics and system properties are all considered. The proposed method includes data pre-processing based on a noise-assisted multivariate empirical mode decomposition method, a two-stage feature selection strategy and remaining useful life prediction based on the trajectory similarity method. An experimental case study is given to evaluate the proposed method. The results show that the systematic prognostic method could be applied in the industrial field.