Net Load Forecasting Based on Residual Attention CNN-BiLSTM Model

Net load forecasting is a crucial task in power system operation and energy management. Accurate prediction of net load helps power companies optimize their generation plans, enhance energy utilization efficiency, and ensure a balance between power supply and demand. To improve the accuracy and precision of net load forecasting, this study proposes a solution based on the Residual Attention CNN-BiLSTM (RA-CNN-BiLSTM) model. The proposed model combines Convolutional Neural Network (CNN) and Bidirectional Long Short-Term Memory (BiLSTM) and incorporates the residual attention mechanism to capture important features and temporal relationships in net load forecasting. Experimental comparisons and evaluations demonstrate that the RA-CNNBiLSTM model achieves lower prediction errors in net load forecasting and outperforms other models significantly.