Özet:
The importance of rectangular porous fins for the transformation of heat through the system is well-recognized to analyze the physical characteristics of material in practical applications. In this study, a neuro-computing based stochastic numerical paradigm has been designed to study the dynamics of temperature distribution in porous fin model by exploiting the strength of artificial neural network (ANN) modeling integrated with global search exploration with genetic algorithms (GAs) and efficient local search with interior-point technique (IPT). The governing porous fin equation is transformed into an equivalent nonlinear second order ordinary differential equation. Effect of heat on rectangular type fin with thermal conductivity and temperature dependent internal heat generation is measured through stochastic solver based on ANN optimized with GA-IPT in case of two different materials, Silicon nitride Si3N4 and Aluminium Al. The proposed technique ANN-GA-IPT has been applied on transformed equation for multi-times and the accuracy, convergence and robustness of designed model has been validated by analysis of variance test.