Long-time wind power prediction based on DCA-TCN-Informer

Due to the complexity and instability of wind power data, the prediction model Informer of previous years often still has insufficient accuracy when dealing with long time wind power prediction. In order to solve this problem, this paper proposes a wind power prediction method based on DCA-TCN-Informer. By combining deep convolutional attention (DCA) and temporal convolutional network (TCN), the method effectively captures the long and short-term dependencies in wind power data, and not only performs well in the overall trend prediction, but also shows higher accuracy in the prediction precision. Through a series of experiments on publicly available wind power datasets, the results validate that the model significantly improves the accuracy of wind power prediction and is robust when dealing with complex nonlinear data. This result provides a new idea to improve the reliability of wind power prediction, which is of great significance to the development of prediction technology in the field of renewable energy.