Reliable estimation of solar irradiance is required for many solar energy applications such as photovoltaics, water heating, cooking, solar microgrids, etc. Deep Learning techniques have shown outstanding behaviour for analysing complex datasets efficiently with high accuracy. Multi-Layer Perceptron, Long-Short Term Memory, and Gated Recurrent Unit techniques are found to be the most competitive techniques in the literature for solar irradiance forecasting. Therefore, in this study, a comparative analysis of those models is carried out using eleven years of NASA satellite data for training and testing. The grid search technique is used to optimize the networks architectures to ensure the best performance of the models for forecasting daily global solar irradiance. The results show that all models have almost the same accuracy with a mean square error close to 0.017 kWh/m2/day. However, the speed of training and the complexity differ for each model. The MLP is found to be the most efficient model due to using a low number of parameters. The study is of importance for reliable solar irradiance forecasting for any location worldwide.
Solar photovoltaic (PV) panel parameter estimation is vital to manage solar-based microgrid operations, for which several techniques have been developed. Solar cell modeling using metaheuristic algorithms is found to be one of the accurate techniques. However, it requires experimental datasets, which may not be available for most of the industrial modules. Therefore, this study proposed a new model to estimate the solar parameters for two types of PV panels using manufacturer datasheets only. In addition, two optimization techniques called particle swarm optimization (PSO) and genetic algorithm (GA) were also investigated for solving this problem. The predicted results showed that GA is more accurate than PSO, but PSO is faster. The new model was tested under different solar radiation conditions and found to be accurate under all conditions, with an error which varied between 7.6212 × 10−4 under standard testing conditions and 0.0032 at 200 W/m2 solar radiation. Further comparison of the proposed method with other methods in the literature showed its capability to compete with other models despite not using experimental datasets. The study is of significance for the sustainable energy management of newly established commercial PV micro grids.