A Novel Hybrid Deep Learning Model For Short-Term Wind Speed Forecasting Based On LSTM And TCN

In recent years,the problems of environmental pollution and global warming have become more and more serious.As people pay more attention to this prob-lem,the proportion of clean energy is increase,and wind energy,as one of the most popular clean energy at present,is widely used all over the world but wind speed prediction is an urgent problem to be solved.Accurate wind speed forecasting can provide strong support for wind farm planning and site selection.Wind speed fluctu-ation has great randomness and unpredictability,which makes difficult to establish a satisfactory model.With the development of data mining and artificial intelligence,wind speed prediction is becoming more and more accurate.This paper presents a new deep learning hybrid model.The model is based on Singular Spectrum Analysis(SSA),Complete Ensemble Empirical Mode Decomposition with Adaptive Noise(CEEM-DAN),Long Short Term Memory(LSTM),Temporal Convolutional Networks(TC-N)and Genetic Algorithm(GA).In this model,SSA and CEEMDAN were used to decompose the original sequence twice,and LSTM and TCN were used to predict the decomposed sequence.The genetic algorithm was used to optimize the global super parameters to get the optimal model.According to the performance of the model in the two data sets,the model proposed in this paper is an effective short-term wind speed prediction model,which can significantly improve the prediction accuracy.