One of the most important and promising types of non-traditional (renewable) energy sources is solar energy. Solar energy has two important advantages: first, its quantity is huge, and it is almost inexhaustible, and second, its use does not hurt the environment. However, the practical use of solar energy is difficult due to the low surface density of solar radiation. In addition, there are problems associated with the solar radiation fluctuations varying during daytime and period of the year. This work aimed to develop a method that allows predicting the output power of the solar panel based on a stochastic model. The proposed stochastic approach considers the following factors: fluctuations (random deviations) of solar radiation at different time intervals, the variable nature of the consumers' load, and the temperature mode of solar photovoltaic panels leading to assess their efficiency. The study presents a theoretical model with the future practical implementation.