Abstract : As Corsica is a non-interconnected island, its energy supply is very special case. Indeed, as all islands, a large part of the electricity production must be generated locally. Often, renewable energies are considered as a good solution to overcome the isolation problem. However, because of their intermittent nature, they are included in a limited way in power systems. Thus, it's necessary to use in addition other energy productions, with main problem the management of the dispatch between these two energy types. This study is related to the solar and PV prediction in order to quantify available energy and to allow the optimal transition between intermittent and conventional energies sources. Throughout this work, we tested different techniques of prediction concerning four horizons interesting the power manager: d+1; h+24, h+1 and m+5. After all these manipulations, we can conclude that according the considered horizon, the prioritization of the different predictors varies. Note that for the d+1 horizon, it is interesting to use an approach based on neural network being careful to make stationary the time series, and to use exogenous variables. For the h+1 horizon, a hybrid methodology combining the robustness of the autoregressive models and the non-linearity of the connectionist models provides satisfactory results. For the h+24 case, neural networks with multiple outputs give very good results. About the m+5 horizon, our conclusions are different. Thus, even if neural networks are the most effective, the simplicity and the relatively good results shown by the persistence-based approach, lead us to recommend it. All the proposed methodologies and results are complementary to the prediction studies available in the literature. In conclusion, we can say that methodologies developed could eventually be included as prediction tools in the global command - control systems of energy sources.