Evolutionary Computing for the Radiative–Convective Heat Transfer of a Wetted Wavy Fin Using a Genetic Algorithm-Based Neural Network

• Published in: Biomimetics (Basel, Switzerland) Vol. 8; no. 8; p. 574
• Main Authors: Poornima, B. S., Sarris, Ioannis E., Chandan, K., Nagaraja, K.V., Kumar, R. S. Varun, Ben Ahmed, Samia
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 01.12.2023
• Subjects: Algorithms; Alternative energy sources; artificial neural network; Cooling; Energy efficiency; Engineering; genetic algorithm; Genetic algorithms; Heat exchangers; Heat transfer; Investigations; Nanomaterials; Natural selection; Neural networks; Nuclear reactors; Radiation; Thermal energy; wavy fin; wet fin; India; genetic algorithm; wet fin; artificial neural network; wavy fin; heat transfer
• ISSN: 2313-7673; 2313-7673
• DOI: 10.3390/biomimetics8080574

Evolutionary algorithms are a large class of optimization techniques inspired by the ideas of natural selection, and can be employed to address challenging problems. These algorithms iteratively evolve populations using crossover, which combines genetic information from two parent solutions, and mutation, which adds random changes. This iterative process tends to produce effective solutions. Inspired by this, the current study presents the results of thermal variation on the surface of a wetted wavy fin using a genetic algorithm in the context of parameter estimation for artificial neural network models. The physical features of convective and radiative heat transfer during wet surface conditions are also considered to develop the model. The highly nonlinear governing ordinary differential equation of the proposed fin problem is transmuted into a dimensionless equation. The graphical outcomes of the aspects of the thermal profile are demonstrated for specific non-dimensional variables. The primary observation of the current study is a decrease in temperature profile with a rise in wet parameters and convective-conductive parameters. The implemented genetic algorithm offers a powerful optimization technique that can effectively tune the parameters of the artificial neural network, leading to an enhanced predictive accuracy and convergence with the numerically obtained solution.