For enhancing the management of energy resources as well as for dependable integration of renewable energy systems to power grids, precise forecasting of solar energy generation is essential. Standard forecasting methods usually do not cope well with the nonlinear functions, fluctuations in the system due to weather changes, and dynamics of solar irradiance. The development of this paper is based on a hybrid forecasting strategy, which aims to improve prediction accuracy by incorporating meteorological information with machine learning techniques. The proposed methodology utilizes weather parameters such as temperature, humidity, cloud cover, and solar irradiance along with Random Forest and Long Short-Term Memory (LSTM) networks. Evaluation on real-world datasets shows that the proposed hybrid model outperformed standalone ones and baseline methods on multiple forecasting performance measures. MAE, RMSE, and R² score measurement proved that the hybrid approach not only decreases the error values but also enhances performance for both short-term and long-term forecasting. The results of this study reveal that using weather data fused with machine learning can efficiently and reliably address the problem of forecasting solar energy.
 

LSTM, Machine learning, Weather, Hybrid data, Solar, Green energy