International Research Journal of Education and Technology

The demand for accurate short-term solar power forecasting has increased with the growing integration of solar energy into the power grid. Predicting solar power generation over short time horizons, such as hourly or sub-hourly intervals, is critical for optimizing grid operations and ensuring energy balance. This study proposes a forecasting model based on Recurrent Neural Networks (RNNs), specifically Long Short-Term Memory (LSTM) networks, to predict short-term solar power output. LSTMs are particularly suited for time-series data due to their ability to capture temporal dependencies. In this work, historical solar power generation data, along with meteorological variables such as temperature, humidity, and solar radiation, are used to train the model. The performance of the proposed RNN model is evaluated against traditional machine learning approaches, demonstrating its superior accuracy in short-term forecasting. Results indicate that the RNN-based model provides reliable solar power predictions, which can be utilized to improve grid management, integrate solar energy more effectively, and optimize energy dispatch in renewable power systems.

Key Words: hort-term forecasting, Solar power, Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), Time-series prediction, Renewable energy, Grid integration, Meteorological data, Energy optimization, Machine learning.