Prediction of Icing Fault of Wind Turbine Blades Based on Deep Learning

Wind turbine generator blade icing prediction is a major technical problem in the field of wind power. Accurate prediction of icing status has important practical significance for ensuring safe operation of wind turbines, reducing operational risks and avoiding loss of power generation. This paper proposes a new method of icing fault prediction based on the combination of deep learning and physical mechanism. According to the real-time data of the wind turbine operation provided by the system, firstly, the data is preprocessed for the height imbalance of the sample class. The decision tree is further used for feature screening, and combined with the physical mechanism of blade icing formation, the features most relevant to icing fault are extracted. Finally, the icing prediction model is established by Convolutional Neural Networks. The accuracy and effectiveness of the model are verified by experiments.