Characteristic signal based on the combination of empirical mode decomposition method and time series AR model Extraction method

• Published in: Journal of physics. Conference series 2021-08, Vol.1986 (1), p.12126
• Main Authors: Bin-Bin, Li, Xiang-Dong, Li, Xu-Yang, Pu, Zhong-Tao, Kang, Zong-Yang, Li, Hui-Zhuo, Zhang, Chen-Xi, Chen, Yue, Hu
• Format: Article
• Language: English
• Subjects: Autoregressive models; Autoregressive processes; Basis functions; Decomposition; Empirical analysis; Fault diagnosis; Roller bearings; Signal processing
• ISSN: 1742-6588; 1742-6596
• DOI: 10.1088/1742-6596/1986/1/012126

In the process of signal decomposition by wavelet theory, the wavelet basis function is artificially selected based on experience, and the method based on empirical mode decomposition is decomposed according to the time scale of the signal itself. The article uses two methods to decompose a certain segmented frequency conversion signal to obtain the intrinsic modal component matrix and the wavelet decomposition coefficient matrix, then calculates the Hilbert time spectrum of the two decomposition matrices. cThe calculations shows that the false information generated by the empirical mode decomposition signal is obviouly more less. Therefore, the empirical mode decomposition method is used to decompose the bearing vibration signal, and the stationary natural mode function obtained is very suitable for establishing an autoregressive AR model to extract the power spectrum of each component for analysis. Finally, the characteristic frequencies of different states of rolling bearings are extracted, which provides support for data-driven fault diagnosis.