Modelling of a flat-plate solar collector using artificial neural networks for different working fluid (water) flow rates

Abstract

The operation of a flat-plate solar collector using three different working fluid flows (water, i.e. 1, 1.6, 2 L/min) has been modelled using the artificial neural networks (ANNs) of computational intelligence technique. The ANNs model has been built at the entrance to predict the outlet temperature in the flat-plate solar collector using measured data of solar irradiance, ambient temperature, inlet temperature and working fluid flow. The results obtained conclude the method is accurate with the three flow rates of the working fluid (water) (e.g. RMSE = 0.1781 °C and R2 = 0.9991 for an ANN prediction of the outlet temperature of the working fluid with 2 L/min test and RMSE = 0.0090 [0,1] and R2 = 0.7443 as a performance prediction test of 1 L/min), flexible when choosing the variables used and easy to apply to any solar collector. The Hottel-Whillier-Bliss (HWB) and the international standard ISO 9806 solar collector models are also described and applied using the data obtained in the tests performed on the flat-plate solar collector. The deviation that occurs with the three different flows of the working fluid (water) used, have been verified and also their repercussion when they are applied in the f-chart method.