Leveraging artificial neural networks approach for thermal conductivity evaluation in porous rectangular wetted fins filled with ternary hybrid nanofluid

The present work deals the temperature transmission of wetted rectangular porous fins of fixed length with an adiabatic tip under Local Thermal Non-Equilibrium (LTNE) model is analysed with the influence of convection and radiation effects. By using the Darcy model and Boussinesq's approximation, the impacts of buoyancy force are considered to estimate the penetration speed within the permeable material. Two energy equations (Solid and Fluid state) are derived for the mathematical model. The fluid state consists of ternary hybrid nanofluid with a combination of $Mo{S}_2+F{e}_3{O}_4+NiZnF{e}_2{O}_4$ nanoparticles with methanol as a base fluid. In addition, both equations are converted to dimensionless non-linear ordinary differential equations by using dimensionless variables, and these equations are solved by using Runge Kutta Fehlberg fourth fifth-order (RKF 45). Further, the average Nusselt number is analysed using an Artificial neural network by applying the Levenberg Marquart backpropagations algorithm. By using this algorithm, the regression analysis, mean square error, and error histogram of the neural network are analysed. In this model, three distinct types of samples are examined, comprising 80% of data points allocated for training the neural network, 10% for testing, and 10% for validation of the artificial neural network (ANN) model. The supremacy of essential aspects of the temperature profile and average Nusselt number is displayed through graphs. However, it is noticed from the results that the surface-ambient radiation parameter levels are decreased and the temperature profile of both solid and ternary nanofluid phase is augmented. The regression coefficient value obtained from ANN model is $R=1$ f, which means the parameters are in strong correlation with each other.