Variational filtering differential strategy with amplitude modulation for fault feature extraction

Abstract Spectral amplitude modulation (SAM) can effectively enhance the features of different signal components due to its superior spectral line editing capabilities. However, noise and interference cannot be separated by the SAM equalization effect. Therefore, this paper proposes a method named variable differential spectrum amplitude modulation to optimize the amplitude modulation process. By designing differential SAM assisted by the modulation Gini index, the optimal differential demodulation order is selected, and bearing fault information within the frequency band is enhanced by frequency correlation. Constructing a variational filtering framework based on spectral envelope and an adaptive resonant band identification method can increase the accuracy of feature extraction under multi-band unknown strong interference. The feasibility of the proposed method is verified using simulation data and experimental signals, with comparisons against other widely used methods confirming its prominent superiority.