Research on the mechanical fault diagnosis method based on adaptive Fourier decomposition and synchronous extraction transform

DOI:	10.11809/bqzbgcxb2023.04.020
Keywords:	synchronous extraction transformation (SET); adaptive Fourier decomposition (AFD); fault diagnosis; frequency aliasing; time frequency analysis
Abstract:	When the traditional synchronous extraction transform (SET) method is used for mechanical fault diagnosis, it is required that instantaneous frequency difference of each adjacent mode of multi component signals is greater than twice of the frequency support range of the selected SET window function. Otherwise, frequency aliasing occurs easily, and the actual signals often cannot meet the condition. In addition, the signals are affected by noise, and instantaneous frequency of the fault is hardly obtained. In view of the shortcomings of the traditional synchronous extraction transform in mechanical fault diagnosis, this paper proposes a mechanical fault diagnosis method based on AFD SET by combining the advantages of adaptive Fourier decomposition (AFD) and SET. The proposed method effectively processes non stationary signals with similar frequencies, accurately expresses instantaneous frequency of the signals, and has a fast convergence speed. The simulation results show that the method solves frequency aliasing in the traditional SET method, and has a higher time frequency concentration. Finally, the proposed method is applied to fault diagnosis of rolling bearing. The diagnosis results further verify the correctness of the proposed method, which can effectively extract the frequency characteristics of fault signals.
Issue:	Vol. 44 No. 4 (2023)
Published:	2023-04-28
PDF HTML