As one of the key components of transmission and support in rotating machinery, rolling bearings will directly affect the operating status of the equipment. Therefore, scholars have proposed various methods to process its fault signals. Among them, spectral amplitude modulation (SAM) is a nonlinear filtering method that can effectively extract bearing fault characteristics. However, due to the high-intensity noise interference in the working environment of the equipment, the interference component dominates the monitoring signal, and the fault characteristics are no longer obvious or even undetectable. To solve the above problems, this paper proposes a local W transform spectral amplitude (LWTSAM) modulation method. First, the method divides the original signal into multiple narrowband signals in different frequency bands and selects the narrowband signal with the most abundant fault information among them; then, the selected narrowband signal uses a windowed Fourier transform (WFT) to obtain the amplitude in the time-frequency domain, and uses different weight indices (MO) to correct the amplitude; Finally, the modified amplitude is combined with its original phase to perform inverse window Fourier transform to obtain the modified signal and its square envelope, and the square envelope under the optimal weight is calculated using unbiased autocorrelation and information entropy to complete the local W transform spectrum amplitude modulation. In this paper, this method is verified through fault data sets. The research results show that this method can effectively reduce the interference of noise on fault diagnosis, and the fault characteristic information obtained is clearer. Compared with SAM method, Autogram method and fast spectral kurtosis diagram method, the superiority of this method is proved.
Ma et al. (Sat,) studied this question.