Research on Electric Power Load Forecasting Based on SO-CatBoost

In the field of electric power load forecasting, traditional forecasting methods face challenges of insufficient data processing capability and difficulties in model optimization. A new method combining the CatBoost algorithm with the Snake Optimization Algorithm (SO) is proposed for power load forecasting. The SO algorithm is used to optimize the hyperparameter configuration of CatBoost to enhance the model's predictive accuracy and generalization ability. The forecast using power data from a specific region in China shows that the SO-CatBoost model significantly outperforms traditional forecasting methods, with RMSE, MAE, and MAPE values of 0.86, 0.66, and 10.78%, respectively. This method provides an efficient and reliable decision-support tool for power system load forecasting, offering significant practical value.