Enhancing PVT/air system performance with impinging jet and porous media: A computational approach with machine learning predictions

Thermal photovoltaic systems (PVT) absorb the sun's energy and convert it into electricity. Increasing the solar cell temperature reduces its efficiency. In the present research, the porous medium and impinging air jet is used for PV cooling to lessen the solar cell temperature. Different arrangements of impinging jet systems (single and multiple) have been considered to evaluate the electrical efficiency of PV. The effects of Re , solar radiation intensity , porosity coefficient, Darcy number , dimensionless porous layer thickness, jet and injection location, jet velocity and angle on PVT efficiency have been inspected. The results show that the porous medium has a positive effect on reducing the PV temperature and increasing the electrical efficiency of PV by increasing the effective thermal conductivity coefficient and reducing the convection resistance. The effect of the characteristics of the porous medium on the electrical efficiency of PV is influenced by porous layer thickness. Porous medium with variable porosity can improve the electrical efficiency of PV up to 6 % compared to uniform porosity. Impingement jet in single and multiple arrangement between 3 and 26 % to improve the electrical efficiency of PV. The impinging jet has a high potential in removing heat from PV. In the integration of PV with multiple jets, more efficiency can be obtained than the single jet mode. Also, by using the data of this survey and the adaptive Neuro-fuzzy inference system (ANFIS) model and Gaussian process regression (GPR) from machine learning algorithms , the electrical efficiency has been estimated and the GPR method has been able to approximate the electrical efficiency well. • The combination of jets and porous foam gradient resulted in an improvement of electrical efficiency and thermal efficiency. • Impingement jet in both single and multiple arrangements improved PVT efficiency • Impingement jet has more uniform temperature distribution on the PV surface • Machine learning algorithms such as ANFIS and GPR were used to estimate electrical efficiency • Using data from this survey and the GPR method provided good approximations of electrical efficiency.