The Application Research of Deep Learning Based Method for Short-term Wind Speed Forecasting

Short-term wind speed forecasting is a difficult work for the randomness of wind speed variation. With the advantage of deep learning, a novel model is proposed to improve the performance of short-term wind speed forecasting. Different from the normal deep learning model, the framework of this new model is composed of convolutional neural network (CNN) and long short term memory (LSTM), extracting common features on high-dimensional time series data at different levels with a more stable prediction effect by CNN and utilizing the context information on the time axis based on the excellent memory ability of the time series data of LSTM. Meanwhile, for the purpose of accelerating calculation and mining the features of data better, the inception structure is also introduced. At last, the experiment results are presented and it shows the improvement of the proposed method, compared to the basic LSTM model.