Two-Stage Short-Term Power Load Forecasting Method Considering Error Compensation
Aiming at the uncertainty and volatility problems in short-term power load forecasting,a two-stage short-term power load forecasting method considering error compensation was proposed.In the first stage,variational mode decomposition was used to decompose the power load data into several simple modes,and the Sparrow Search Algorithm based on firefly disturbance optimization was used to optimize the hyperparameters of the two-way long short-term memory neural network,and a load forecasting model was established to obtain the initial load forecasting power values.In the second stage,comprehensively considering the error series and external influencing factors,error compensation model was established and the error compensation value was obtained,and the values of the two stages were added together to determine the final load prediction result.Compared with other combined prediction methods using the actual load data of the two regional cells to simulate the example,results show that the method proposed in this study has higher prediction accuracy where the average absolute percentage error and root mean square error can respectively reach 1.26% and 29.76 kW,thus the effectiveness of the proposed method was verified.
万方数据
维普
