Application of neural networks for predicting the energy yield of a photovoltaic module
Zastosowanie sieci neuronowych do predykcji uzysku energetycznego z modułu fotowoltaicznego
Łukasz Dyoniziak, Joanna Aleksiejuk-Gawron
Streszczenie
The article presents the application of artificial neural networks for predicting the energy yield of a monocrystalline
photovoltaic module. The research was based on measurement data from a meteorological station and a photovoltaic systems
laboratory, located at the Institute of Mechanical Engineering, Warsaw University of Life Sciences. The input data included
the sum of solar radiation, external temperature, and module temperature, while the model's output was energy yield.
MATLAB and the Neural Net Pattern Recognition tool were used to develop the predictive model. The results demonstrated
high predictive accuracy, confirmed by, among other things, low mean square error values and high correlation coefficients,
suggesting the potential of utilizing artificial neural networks for optimal energy yield management in photovoltaic systems.
photovoltaic module. The research was based on measurement data from a meteorological station and a photovoltaic systems
laboratory, located at the Institute of Mechanical Engineering, Warsaw University of Life Sciences. The input data included
the sum of solar radiation, external temperature, and module temperature, while the model's output was energy yield.
MATLAB and the Neural Net Pattern Recognition tool were used to develop the predictive model. The results demonstrated
high predictive accuracy, confirmed by, among other things, low mean square error values and high correlation coefficients,
suggesting the potential of utilizing artificial neural networks for optimal energy yield management in photovoltaic systems.