Effective wind speed estimation study of the wind turbine based on deep learning

Wind speed is the driver of wind turbines, and the precise estimate of that makes it possible to improve the control effects and the efficiency of energy production. This paper proposes a method of effective wind speed estimator that considers the variation in blade radius and reconstructs the mapping of the aerodynamic. First, the proposed method utilizes available data of the wind turbine to train two neural network models based on radial basis functions (RBF). The models estimate the effective radius and reconstruct the aerodynamic mapping surface, respectively. Then, on this basis, train a wind speed estimation model based on the Long Short-Term Memory (LSTM) neural network, which can effectively estimate the current wind speed in real-time and predict the wind speed of the next time step as the reference for the following estimation. In addition, the RBF model and LSTM model can be updated and improved adaptively based on new data to ensure the accuracy of an effective wind speed estimator. Finally, the proposed wind speed estimation method is compared with the existing methods. The experimental results show that the proposed method has strong anti-interference characteristics, and improves the effective wind speed estimation accuracy over 70% on average.(c) 2022 Elsevier Ltd. All rights reserved.

Automated summary

This paper proposes an effective wind speed estimation method for wind turbines, which considers the variation in blade radius and reconstructs the aerodynamic mapping. The method utilizes neural network models to estimate and reconstruct wind speed, effective radius, and aerodynamic mapping, and employs LSTM neural network for real-time estimation and prediction. Experimental results demonstrate that the proposed method has high accuracy and anti-interference capability.