Abnormality Detection Method for Wind Turbine Bearings Based on CNN-LSTM

Abstract

Wind turbine energy generators operate in a variety of environments and often under harsh operational conditions, which can result in the mechanical failure of wind turbines. In order to ensure the efficient operation of wind turbines, the detection of any abnormality in the mechanics is particularly important. In this paper, a method for detecting abnormalities in the bearings of wind turbine energy generators, based on the cascade deep learning model, is proposed. First, data on the mechanics of wind turbine generators were collected, and the correlation between the data was studied in order to select the parameters related to the bearing temperature. Then, the logical relationship between the observation parameters and the target parameters was established based on a one-dimensional convolutional neural network (CNN) and a long short-term memory (LSTM) network, and the difference between the predicted temperature and the actual temperature was assessed using the root mean square error evaluation model. Finally, a numerical example was used to verify the operational data from a wind farm unit in northwest China. The results show that the CNN-LSTM model proposed in this paper can detect abnormalities earlier in the state of the main bearing than the LSTM model, and the CNN-LSTM model can detect abnormalities in the main bearing that the LSTM network cannot find.

Data-Driven Models Applied to Predictive and Prescriptive Maintenance of Wind Turbine: A Systematic Review of Approaches Based on Failure Detection, Diagnosis, and Prognosis

“…Ref. [74] proposed a hybrid model composed of CNN and LSTM networks to detect abnormalities in the bearings of wind turbine generators based on deep learning. The results show that the CNN-LSTM model was able to detect abnormalities in the main bearing state earlier than the LSTM model.…”

Fault Diagnosis of Gearbox Based on CNN and LSTM with Attention Mechanism

“…CNN is a kind of deep feedforward neural network with weight sharing and local connectivity features [13], which has very powerful deep feature extraction and pattern recognition capabilities, and has been widely used in many fields such as image recognition. A typical convolutional neural network mainly consists of a convolutional layer, a pooling layer and a fully connected layer.…”

The Prediction of the Remaining Useful Life of Rotating Machinery Based on an Adaptive Maximum Second-Order Cyclostationarity Blind Deconvolution and a Convolutional LSTM Autoencoder

“…To improve the quality of the LSTM, a new model integrates the advantages of CNN and LSTM to address the limits for the RUL prediction of rolling bearings. This method can extract sensing data to monitor health states, preserve these benefits, overcome the overfitting of spatial fluctuations, and achieve efficient and accurate health monitoring [ 17 , 18 , 19 , 20 ]. In addition, some autoencoder theories are applied to the CNN and LSTM to improve the prediction performance [ 21 , 22 , 23 , 24 ].…”