Jianzhou Wang

In the past few decades, wind power generation has gradually become an important source in the global energy market, which effectively meets the energy demand of human production activities. However, the instability and diffusion of wind speed bring difficulties to the wind energy development and promotion. For improving the predictive accuracy of original sequence, scholars have proposed a variety of prediction models, but many current forecasting models often neglect the importance of data processing and can be easily limited by a single model, which causes poor performances. Therefore, a combined model is built that mainly includes the complete ensemble empirical mode decomposition with adaptive noise, several single models, and multi-objective ant lion optimization algorithm. This combined model not only reduces the impact of high-frequency noise, but also extracts original sequences features as much as possible, combines the advantages of multiple single models, and greatly improves forecasting accuracy and stability.