Medium and Long-Term Power Load Forecasting Based on Feature Extraction and CNN-Bi-GRU-AM

Medium and long-term power load forecasting with days as the forecast step length is of great significance to power grid planning, power supply configuration and power grid stability analysis. Aiming at the difficulties faced by medium and long-term load forecasting, such as long-time span, strong volatility and many influencing factors, this paper proposes a medium and long-term power load forecasting method CNN-Bi-GRU-AM based on feature extraction, which uses artificial intelligence algorithms to capture and highlight the long and short-term features of historical data to predict the power load in the target year. The simulation results show that this method reduces the error by 37.68%, 39.25%, 43.63%, and 50.17% compared with the basic models. This method can effectively improve the accuracy of medium and long-term power load forecasting and provide data reference for power system planning and operation.