Rolling bearings are key parts of rotating machinery and their failure will lead to equipment failure. Therefore, it is very necessary to extract fault characteristics of rolling bearings. When there are complex interference frequencies in the rolling bearing signals, the fault characteristic signals will be difficult to identify. In order to solve the above problems, this paper combines the wavelet threshold denoising and the spectral amplitude modulation (SAM) algorithm to apply to the low-speed bearing fault diagnosis. Firstly, perform wavelet thresholding denoising on the raw signals to obtain denoising signals. Secondly, the modified signals are obtained by SAM of the denoising signals. Finally, the fault features are extracted by envelope analysis of the modified signal. The proposed method is applied to experimental signals. Experimental results show the effectiveness of the proposed method in low-speed bearing fault diagnosis

fault diagnosis, SAM, the wavelet threshold denoising, Low-speed, Feature extraction