Predicting Electrical Load Demands Using Neural Prophet-Based Forecasting Model

The appropriate utilization of electrical load forecasting should be a top priority to accomplish the emission reduction targets and achieve the ideal balance between power generation and consumption. Load forecasting is an extremely critical task for the utilities because it aids in the efficient planning and optimal utilization of the resources. This results in less wastage, maintenance of equipment and reduction in negative environmental impacts. This paper presented a novel strategy of utilizing a neural prophet-based load forecasting model for prediction electrical load demand. Facebook (Fb) has designed neural prophet model based on its successful Fb-prophet model. The neural prophet-based model has been chosen due to its qualities in terms of explainability, scalability, flexibility and user-friendliness. The neural prophet-based methodology extensively outperformed in terms of prediction accuracy by 98.97% over other state-of-the-art models. The root mean square error (RMSE) has been considerably reduced to 101.4592 and achieves a mean absolute percentage error (MAPE) of 1.03%. The proposed neural prophet-based model produced lowest deviation of MAE. The achieved results are at least 1.638% more accurate than other standard models.