Fig. 3: The framework flow of this study.

This figure depicts the specific workflow of this study in detail. Initially, the first module outlines the steps involved in data acquisition and preprocessing. Subsequently, the second module delineates the construction of the forecasting model, along with the computation of forecasting errors based on shape and temporal criteria. The third module showcases the optimization process for the learnable parameters within the model. Finally, the fourth module enumerates the evaluation system and the statistical tests conducted to assess the model’s forecasting performance.