Power load demand forecasting is an important part of power operation. Due to the influence of various factors such as user side and weather on power load, there is a problem of low accuracy in current medium and long-term power load demand forecasting. Therefore, a medium and long-term forecasting algorithm considering user side under the dual carbon background is proposed. Fuzzy clustering method is used to obtain cluster power load curves for processing of user side data. The missing values of the influencing factor data are filled in through Langrange interpolation method, and the main influencing factor data is selected through grey correlation analysis method. The cluster power load curve and main influencing factor data

inputted into the attention mechanism long short-term memory network model to achieve load demand prediction. The experimental results show that the proposed method has higher accuracy in load demand prediction and better practical application effect.

power load curve, fuzzy clustering method, medium-and long-term demand forecasting, langrange interpolation method