Small Fault Diagnosis of Front-end Speed Controlled Wind Generator Based on Deep Learning

In view of the difficulty in diagnosing the early small faults of front-end controlled wind generator (FSCWG), this paper proposes a small fault diagnosis methods based on deep learning. The method adopts a deep learning method, uses vibration data under several different small fault patterns of FSCWG as input of the model and gets deep learning diagnosis model by learning complicated implicit layer structure and training. Then using the trained network to extract feature of FSCWG from original vibration data by layer-wise, and fully excavate the associations among the data and form a more abstract executive property categories or characteristics, to improve the diagnosis accuracy. The results show that compared with the traditional fault diagnosis method of neural network (NN) and support vector machine (SVM) method, the small fault diagnosis method based on deep learning enhances the small fault diagnosis accuracy in the process of generator operation. Key-Words: front-end speed controlled wind generator; small fault; deep learning; convolutional neural network