Fault Diagnosis Method for Rolling Bearings Based on EEMD and PCA

Based on the vibration generation mechanisms of rolling bearings, a mathematical model is established to simulate their vibration signals under various fault conditions. Modern signal processing techniques are applied to the fault diagnosis of rolling bearings, with simulated data serving as the research object. Targeting the nonlinear and non-stationary characteristics of bearing vibration signals, the following methodology is implemented: First, the vibration signals are decomposed using Ensemble Empirical Mode Decomposition (EEMD). A sensitivity evaluation algorithm is then applied to assess all intrinsic mode functions (IMFs), from which the IMFs sensitive to fault features are selected to reconstruct the vibration signals. Features in the time domain, frequency domain, and wavelet packet decomposition are extracted from the reconstructed signals. Principal Component Analysis (PCA) is employed to process these features, and fault identification in rolling bearings is performed using the first, second, and third principal components in both two-dimensional and three-dimensional spaces. The results demonstrate that the proposed method achieves a fault diagnosis accuracy exceeding 96.6% for rolling bearings.