Solar power grid integration has increased tremendously in the global electricity market. However, further increase in solar power grid integration has been restricted by the intermittent nature of solar energy supply. For this reason, researchers have developed different mathematical models which could predict the available solar energy radiation and the actual solar photovoltaic energy generated at a given location. Hence, the present study proposes a novel (enhanced multilayer perceptron neural network, MLPNN) model for predicting the daily solar energy generated by a 1.2 MW PV power plant. The stability of the MLPNN model was compared with results obtained from the multiple nonlinear regression (MNR) model and the generalized regression neural network (GRNN) model. The results showed that the enhanced MLPNN model outpaced the MNR and the GRNN models by presenting the lowest normalized root mean square error (nRMSE), the lowest minimum absolute percentage error (MAPE), and the best coefficient of determination (RSQ) for both the rainy [6.09, 5.93, 93.53]% and the dry [6.12, 4.16, 90.77]% seasons, respectively.<
Alan : Eğitim Bilimleri; Fen Bilimleri ve Matematik; Sağlık Bilimleri; Sosyal, Beşeri ve İdari Bilimler
Dergi Türü : Uluslararası
Benzer Makaleler | Yazar | # |
---|
Makale | Yazar | # |
---|