Load Forecasting Method Considering Weather Factors based on LSTM and XGBoost

In order to further improve the accuracy of power load forecasting, this paper proposes a load forecasting model LSTM-XGBoost that considers weather factors. This model first uses the LSTM neural network to extract long-term dependencies from historical data, and then uses the XGBoost model to fit the historical load data and long-term dependencies and make predictions, which can fully consider the information in historical load data and weather data. Experimental results show that compared to LSTM, XGBoost, and GRU, the proposed model improves the forecasting accuracy by an average of about 0.2 percentage points, achieving higher forecasting accuracy and the involvement of weather factors can improve forecasting accuracy to some extent.