摘要：With the rapid consumption of fossil fuels, traditional power generation methods not only cannot issue future energy needs, but also bring serious environmental problems. As a clean and renewable energy, wind energy plays an increasingly important role in energy supply structure. However, the wind speed itself is intermittent, unstable and random, which brings severe challenges to wind power generation. Aimed at improving the accuracy and reliability of short-term wind speed forecasting, this paper proposes a new hybrid model. The model includes time-varying filter, modal decomposition, permutation entropy, adaptive noise modal decomposition, adaptive neuro-fuzzy inference system (ANFIS), packet data processing method neural network (GMDH), and improved monarch butterfly optimization algorithm (IMBO). First, the original wind speed sequence is significantly decomposed twice to obtain the sub-sequence to be predicted. Then, the reconstructed data uses ANFIS and GMDH neural network models to predict sequences in different frequency domains to get prediction results. To further improve the performance of the model, the improved monarch butterfly optimization algorithm is used to modify the model parameters. Finally, the final prediction result is obtained by summing the prediction results of each component. In addition, for verifying the performance of the model, this paper designs six sets of comparative experiments from two dimensions to verify the model on three data sets. The results show that the model proposed in this paper has high prediction accuracy and good stability.
