In order to improve the prediction accuracy of wind power output, an adaptive noise-complete ensemble empirical mode decomposition (CEEMDAN)-Bayesian optimization (BO)-long and short-term time-series network (LSTNet) is proposed to predict the output power of wind turbines in the short term. Firstly, the data are cleaned, and then, CEEMDAN is used to decompose the original power data after cleaning, and several sub-sequences are obtained. The decomposed subsequences are input into the LSTNet model, the LSTNet hyperparameters are optimized by using BO algorithm, and output the prediction results of the subsequences. Finally, the prediction results of each sequence are superimposed and reconstructed to obtain the final prediction results. Through the example simulation of the measured data of a wind farm unit in Weinan, ablation analysis and comparative analysis are set up. The results show that compared with other models, the prediction accuracy of the proposed method is effectively improved.