Resumo

HERRERA CASANOVA, Reinier et al. Intelligent predictive model proposal for a photovoltaic plant. Rev cuba cienc informat [online]. 2022, vol.16, n.1, pp. 144-162.  Epub 01-Mar-2022. ISSN 2227-1899.

The objective of this work is to present the development of a dynamic model structure for the prediction of electricity generation in a photovoltaic plant. Traditionally, a model for the prediction of electricity generation in this type of plant is based on two models, one for the prediction of solar irradiance and a second model to describe the relationship between solar irradiance and generated power. The main climatological variables considered are solar irradiation and ambient temperature, while from the technological point of view, the surface cleanliness of the solar panels is considered, as well as the operating point of the plant, depending on the period of the year and the time of day. The model presented considers solar irradiation and ambient temperature as input variables, while developing the non-linear modeling between irradiation and generated power, considering as disturbances the shading (partial or not) on the modules and the cleaning of the surface of the photovoltaic panels. The work presents a technological description of a plant and the temporal and frequency characterization of real data sets, from which the conception of the structure of the most suitable model for the application of techniques based on artificial intelligence, specifically deep learning, is developed. Finally, the proposed model is used to perform the direct prediction of the generated power based only on historical data obtained in the photovoltaic plant of the Central University "Marta Abreu" of Las Villas. The results obtained for a medium-term prediction horizon and for different times of the year are accurate, which demonstrates the effectiveness of the proposed prediction method.

Palavras-chave : Photovoltaic plant; Photovoltaic plant model; Generated power prediction; Deep learning..

        · resumo em Espanhol     · texto em Espanhol     · Espanhol ( pdf )