Bearing Fault Diagnosis Method for Marine Vessels Based on Empirical Mode Decomposition and Informer Neural Network

Fault diagnosis of marine vessel bearings is crucial for ensuring maritime transportation safety. This study proposes a bearing fault diagnosis method based on Empirical Mode Decomposition and Informer network. First, the research employs Empirical Mode Decomposition to decompose fault signals, enhancing local features and multi-scale time-frequency information within the signals. Next, an Informer neural network is established to perform fault diagnosis based on intrinsic mode components. Finally, the proposed method is validated using the Case Western Reserve University rolling bearing dataset. Experimental results demonstrate that the EMD-Informer approach achieves an overall accuracy of 93.15%, significantly outperforming baseline models.