Optimizing chemical reactive flow of hybrid nanofluid with heat generation and thermal radiation using artificial neural network: Application to Bio-Nanomedicine
Saba Liaqat, Munawar Abbas, Belgacem Bouallègue, Adnan Burhan Rajab, Ibtehal Alazman, Nouf Abdulrahman Alqahtani, Mustafa Bayram, Ilyas Khan
Abstract
In this article, the impact of heat generation and thermal radiation effects on the chemical reactive flow of hybrid nanofluid inside a stenotic artery using artificial neural network are briefly discussed . A hybrid nanofluid including gold and ferric-oxide nanoparticles and blood as the base fluid is used. Understanding the behavior of blood flow in artery disease, especially when plaque accumulation results in stenosis, is helpful in the biomedical sciences as it provides information on targeted drug delivery, thermal therapy, and the management of vascular diseases. These models aid in the technical design of biomedical devices and cooling systems based on nanofluids , increasing their efficiency through the use of optimal mass transport and heat properties. The incorporation of artificial neural network speeds up the prediction of solutions, increasing the viability of real-time therapeutic and diagnostic planning. Using a suitable similarity variable method, partial differential equations have been transformed into dimensionless ODEs (ordinary differential equations). The built-in bvp4c solver in the mathematical program MATLAB is then used to solve the ODEs both numerically and graphically. An Artificial Neural Network integrated with an Intelligent Bayesian Regularization Scheme (ANN-IBRS) is trained using the reference dataset of the bvp4c method's solutions. Regression analysis, MSE (Mean square error) and error histogram evaluations are used to evaluate the neural network's performance. The numerical solutions of the fluid dynamics system are carefully taken into consideration in order to further decrease the MSE. By contrasting setups incorporating MSE values, error histograms, state transitions, correlation metrics, and regression results, the stochastic technique's dependability and effectiveness are demonstrated.