A Fault Diagnosis Method for Wind Turbine Gearboxes Considering Wind Energy Characteristics

The inherent non-stationarity, weak periodicity, and abrupt transient mutations of wind turbine gearbox vibration signals present substantial challenges to traditional diagnostic frameworks. To improve diagnostic accuracy, this paper proposes a fault diagnosis method for wind turbine gearboxes based on the Snake Bird Optimization Algorithm (SBOA)-optimized Variational Mode Decomposition (VMD) and an improved Bidirectional Gated Recurrent Unit (BiGRU). First, SBOA is employed to adaptively optimize the number of modes and the penalty factor in VMD, enhancing the effectiveness of signal decomposition. Second, an improved BiGRU model is constructed, featuring a three-channel BiGRU architecture, along with the introduction of periodic skip connections, a temporal attention mechanism, and a time-decay loss function to strengthen the extraction and fusion of temporal features. Finally, a Softmax classifier is used to achieve precise identification of fault types. Experimental results demonstrate that the proposed method achieves an average accuracy of 99.40% in diagnosing five types of faults, significantly outperforming comparative methods. Ablation studies further validate the effectiveness of each improved module. The proposed method provides reliable technical support for the intelligent operation and maintenance of wind turbine gearboxes.