The integration of renewable energy sources into the power grid is becoming increasingly important as the world moves towards a more sustainable energy future. However, the intermittent nature of renewable energy sources can make it challenging to manage the power grid and ensure a stable supply of electricity. In this paper, we propose a deep learning-based approach for predicting energy demand in a smart power grid, which can improve the integration of renewable energy sources by providing accurate predictions of energy demand. We use long short-term memory networks, which are well-suited for time series data, to capture complex patterns and dependencies in energy demand data. The proposed approach is evaluated using four datasets of historical energy demand data from different energy distribution companies including American Electric Power, Commonwealth Edison, Dayton Power and Light, and Pennsylvania-New Jersey-Maryland Interconnection. The proposed model is also compared with two other state of the art forecasting algorithms namely, Facebook Prophet and Support Vector Regressor. The experimental results show that the proposed REDf model can accurately predict energy demand with a mean absolute error of 1.4%. This approach has the potential to improve the efficiency and stability of the power grid by allowing for better management of the integration of renewable energy sources.