Error Location Analysis of Wind Power Prediction Based on EMD-LSTM

In order to reduce the prediction error of short-term wind power, an EMD-LSTM based short-term wind power combination prediction model is proposed. First, the EMD algorithm is used to decompose the original wind power data into several eigenmode function components and trend components, and then each component is combined with NWP (Numerical weather prediction) wind resource data to establish a long and short term memory network for regression prediction. Finally, the prediction results are compared with the actual wind power to verify the prediction performance of the proposed model. The experimental results can reasonably predict the wind power with high prediction accuracy. At the same time, an error analysis method based on wind power model is proposed. By decoupling method, we find the cause of power prediction error, that is, the error caused by each link. This method can effectively excavate the causes of errors in the prediction model and the proportion of errors in each link, and provide accurate support for further improvement of wind power prediction errors.