Research on the Impact of the Differencing Operator on Ensemble Learning Algorithms in the Case of Peak Load Forecasting

Abstract

Peak load forecasting is a critical aspect of power system operations and planning. Accurate forecasting of peak loads significantly impacts the overall efficiency and reliability of a power system. Among the numerous load forecasting methods that are used, ensemble learning algorithms have emerged as a popular choice due to their high accuracy. In this research, the author proposes an innovative methodology that integrates the Differencing Operator with the Sliding Window procedure for training and predicting peak loads using commonly employed ensemble learning models such as GBDT, XGBoost, LightGBM, and CatBoost. The performance of the proposed approach was evaluated by analyzing the prediction error and execution time. The results obtained demonstrated improved accuracy in peak load forecasting, with no impact on execution time.