Optimization Research on Short-Term Power Load Forecasting Model Combining LSTM and Residual Networks

Short-term power load forecasting is a fundamental aspect of power system scheduling and energy management, as its accuracy directly affects the stability and economic efficiency of grid operations. To improve forecasting accuracy, this paper proposes a short-term power load forecasting model that combines Long Short-Term Memory networks with residual networks. The model leverages LSTM to capture the temporal dependencies in power load data while incorporating a residual network structure to mitigate the vanishing gradient problem in deep network training, thereby enhancing the model's expressive capability and generalization performance. To validate the effectiveness of the proposed model, experiments were conducted using real power load data from the Panama and Malaysia datasets and compared with traditional forecasting models. The experimental results demonstrate that the proposed model significantly outperforms baseline models in terms of both forecasting accuracy and robustness. This study provides an efficient deep learning solution for short-term power load forecasting, offering valuable insights for intelligent power system management and optimized scheduling.