Combining CFD and artificial neural network techniques to predict the thermal performance of all-glass straight evacuated tube solar collector

Highlights: • Experimental research of all-glass straight-through evacuated tube. • Performance prediction of all-glass straight-through evacuated tube (ETC). • Artificial neural networks (ANN) are used for prediction. • Linking simulated tube outlet temperature to ANN improves the prediction accuracy. • Convolutional neural network combined with thermal CFD model provided best accuracy. Thermal performance modelling and performance prediction of a novel all-glass straight-through evacuated tube collector is analyzed here. A mathematical model of the tube was developed and incorporated into CFD software for numerical performance simulation. To improve the thermal performance prediction of the collector, different artificial neural network (ANN) models were considered. A comprehensive experimental dataset with more than 200 samples were employed for testing of the models. Integrating the thermal simulation model with the ANN models by using modelled collector output as one of the input models, significantly improved the prediction accuracy of the ANN models. The predictions based on the CFD model alone gave the poorest accuracy compared to the ANN models. The convolutional neural network (CNN) model proved to be the best ANN model in terms of prediction accuracy.