ABSTRACT
This paper presents a comparative study between artificial neural networks (ANNs) and an adaptive neuro-fuzzy inference system (ANFIS) to estimate the daily electric energy production of a parabolic trough solar thermal power plant (PTSTPP) in Eastern Morocco. The inputs of the ANN and ANFIS models are daily direct normal irradiation (DNI), day of the month, daily average ambient temperature, daily average wind velocity, daily average relative humidity, and previous daily electric production. To select the best architecture, several models were developed. For the ANN model, we changed the number of hidden neurons, while we varied the number of radius for the ANFIS model. Estimated results indicate that the total electric energy accumulated for the validation year was about 39341 MWh and 39907 MWh, respectively, for the ANN and ANFIS models. The difference between predicted and experimental yearly energy production indicates an underestimation of about 9.58% and 10.86%, respectively, for the ANFIS and ANN models.
