In recent years, renewable energy systems have played a central role in energy management. In effect, they are auxiliaries to the conventional power grid. In order to stabilize and regulate energy supply, it is necessary to monitor fluctuations in the energy output of solar power plants. In this study, we adopted a modern artificial intelligence (AI) model to predict Daily Shortwave Downward Radiation using different models. Based on the results achieved in previous studies, we worked on two models CNN model and LSTM. After We using hybrid model CNN-LSTM to benefit the advantage of two models. All are investigated to predict surface shortwave downward radiation based for univariate time series. We using data of the surface shortwave downward radiation for prior years, we forecast the surface shortwave downward radiation at the next time. The model evaluation based on different sets of metrics. The results show that these models were very acceptable, especially if we compared with another classical model. The result obtained from Mean Absolute Error equals 0.3663.