Fault Diagnosis of Wind Turbines Based on TimeGAN-Stacking

The lack of sufficient fault samples to support the training of fault diagnosis models in wind turbines often leads to monitoring systems missing or misreporting faults. In response to the above issues, this article proposes a fault diagnosis method for wind turbine pitch system based on TimeGAN-Stacking. At the data level, due to the imbalance of the original sample categories, a Time Generative Adversarial Network (TimeGAN) is used to track the dynamic changes in the probability distribution of wind turbine operation data step by step, while optimizing the global and local distribution of the generated samples, effectively balancing and expanding the comprehensive sample set of wind turbine faults; At the model level, establish a Stacking integrated model that integrates the advantages of multiple fault diagnosis devices to further improve fault diagnosis capabilities. Finally, the proposed method was tested based on actual wind field data, and the results showed that the TimeGAN-Stacking fault recognition method can effectively diagnose wind turbine faults and has good recognition performance.