关键词: 风电机组, 分形特征, 状态辨识, 特征提取

Abstract: To ensure the safe and reliable operation of wind turbines,a model based on Convolutional Neural Network (CNN) convolution neural network with fractal features for wind turbine operation state recognition was proposed.The initial sample feature data set was established based on fractal and other statistical features.Density-Based Spatial Clustering of Applications with Noise (DBSCAN) unsupervised clustering method was used to mark the sample states,and regularization feature selection method was used to determine the final sample feature data set.The state recognition model was established based on CNN.The experiment results showed that the accuracy of the model was 98.925%,which provided scientific reference and theoretical guidance for “prior maintenance” mode of wind turbines.It could be effectively applied for state recognition of complex electromechanical systems and provide a basis for data mining and pattern recognition of wind turbines and other complex electromechanical systems.

Key words: wind turbine, fractal feature, state recognition, feature extraction