REDf: a deep learning model for short-term load forecasting to facilitate renewable integration and attaining the...

Area of Science:

• Electrical Engineering
• Artificial Intelligence
• Sustainable Energy Systems

Background:

• Integrating renewable energy sources is crucial for global sustainable energy goals (UN SDG 7).
• The intermittency of renewables poses challenges for power grid stability and management (UN SDG 9).
• Accurate energy demand forecasting is essential for efficient grid operation and renewable energy integration.

Purpose of the Study:

• To propose a deep learning model for accurate short-term energy demand prediction in smart power grids.
• To enhance the integration of renewable energy sources by improving demand forecasting.
• To support UN Sustainable Development Goals 7, 9, and 13 through improved grid management.

Main Methods:

• Development of a deep learning model, specifically Long Short-Term Memory (LSTM) networks, for time series forecasting.
• Evaluation of the proposed model using four historical energy demand datasets from major US utility companies.
• Comparison of the model's performance against state-of-the-art algorithms: Facebook Prophet, Support Vector Regression, and Random Forest Regression.

Main Results:

• The proposed REDf model achieved a Mean Absolute Error (MAE) of 1.4% in energy demand prediction.
• The model demonstrated superior accuracy compared to Facebook Prophet, Support Vector Regression, and Random Forest Regression.
• Experimental results confirm the model's capability for accurate short-term energy demand forecasting.

Conclusions:

• The REDf deep learning model offers a highly accurate solution for energy demand prediction in smart grids.
• The model can significantly enhance the stability and efficiency of power grids with high renewable energy penetration.
• This approach effectively supports the achievement of UN SDGs 7, 9, and 13 by facilitating renewable energy integration and climate action.