Short-term load forecasting based on two-stage error compensation

Power load forecasting is related to the optimal dispatch and resource allocation of the power grid, but load forecasting is bound to be accompanied by errors. This paper proposes a short-term load forecasting method based on two-stage error compensation. The data after denoising using the EMD method is used for the training of the prediction model and the error prediction model, and the load forecast value and the forecast error value of the day of the forecast are obtained respectively, the forecast error is used for the first correction of the forecast load. Due to the “peak-valley” characteristics of the electric load, the value of each “peak-valley” points is predicted by the trend extrapolation method for the second correction. Finally, a case analysis was carried out based on the power load data of a certain place in Ningxia, The results show that the prediction accuracy of the proposed method is greatly improved, which verifies the feasibility of the method.