Hybrid Model of LSTM and CNN (2D) for Solar Radiation Prediction

Accurate forecasting of solar radiation is crucial for maximizing the efficiency of solar power plants and maintaining grid stability. However, many existing models struggle to effectively capture both the spatial dependencies among meteorological variables and the temporal patterns influencing solar radiation. A hybrid deep learning model that combines LSTM with CNN networks is suggested as a solution to these drawbacks. The CNN component captures local interactions between factors like temperature, humidity, and solar radiation by extracting spatial patterns from past weather data. By integrating spatial and temporal analysis, the hybrid model overcomes the drawbacks of standalone approaches, yielding superior predictive performance. The model obtains a coefficient of determination (R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>) of 0.9674, MAPE of 5.85%, MAE of 0.1246, and RMSE of 0.1589. This performance demonstrates the model’s potential as a robust tool for maximizing solar radiation forecasting and supporting sustainable energy management.