Hybrid model for electricity price prediction based on SCINet-GConv-CA

Fluctuations in electricity price data are often caused by multiple factors, which increases the complexity of forecasting electricity prices and selecting features. To improve the accuracy and reliability of electricity price predictions, a hybrid model based on SCINet-GConv-CA is proposed. Group convolution is utilized to enhance the feature extraction capability of the model and to reduce its computational complexity. The feature transformation module is then used to improve the overall performance and generalization ability of the model, while mining deeper features of the data. Finally, the channel attention mechanism is utilized to enhance the feature expression capability and improve the prediction accuracy. Experiments are conducted on a regional electricity price dataset from Australia and Spain, and the results show that the prediction performance of the model proposed in this paper is significantly improved compared with comparative models such as SCINet.