A method of convolutional neural network based on frequency segmentation for monitoring the state of wind turbine blades

•A method based on frequency segmentation is proposed based on the energy distribution of wind turbine blades in different frequency bands.•A model was developed that incorporates VMD, Mel spectrogram, and frequency segmentation and integrates these with convolutional neural networks.•Considering that the experimental setting has been subjected to noise reduction processing, it is proposed that Gaussian white noise be manually added to the original data for VMD filtering.•Given the inadequacy of data training in real-life situations, a judgment approach based on score coefficients is provided.•The model was evaluated using experimental data and produced satisfactory results.