Multistep prediction for short-term wind speed based on the MLP and LSTM method with rankpooling

The actual wind speed data suffers from the intermittent and fluctuating property, which implies that it is very difficult to forecast wind speed with high accuracy by applying single or shallow models. Hence, with the purpose of improving the forecasting accuracy and obtain better forecasting results, in this paper, a novel hybrid deep learning model is proposed for multistep forecasting of wind speed, which is intuitively abbreviated as LR-FFT-RP-LSTM and LR-FFT-RP-LSTM. Under these formulated model, the rankpooling method is firstly presented to extract local features of the raw meteorological data, and the Fast Fourier Transformation (FFT) is adopted to extract local and global features of the raw meteorological data to obtain pre-processed data, and the data obtained is then integrated with the original data using the two procedures to produce two input datasets. Then, deep learning model named multi-layer perceptron method (MLP) and long short-term memory (LSTM) are adopted to predict the wind speed dataset. The target prediction results are then obtained by integrating the preliminary prediction findings using the linear regression method.Practical wind speed data from 2010 to 2020 are exploited to evaluate the performance of the proposed model. Case study results indicate that the proposed model for wind speed has a superior forecasting capability. Moreover, the proposed hybrid model is very competitive compared to the state-of-the-art single model and other hybrid models involved in this paper.