An intelligent soft computing model for predicting the thermal behavior of blood-based trihybrid nanofluids flow in biomedical drug delivery applications

In biological studies, nanofluids play a crucial role in targeted drug delivery to specific parts of the human body, cancer treatment via hyperthermia, and other applications. The main innovation of this study is the application of soft computational techniques, specifically the design and implementation of a nonlinear autoregressive exogenous input network backpropagated with the Bayesian Regularization approach (NAEI-BRA) to predict the behavior of blood-based trihybrid nanofluid flow through a porous medium, influenced by magnetization effects, along with the Darcy-Forchheimer inertial drag model (BTHN-DFDM). The physical properties of the metal nanoparticles AA7075 and AA7072 in the base fluid, blood, are used to develop a model describing their interaction with zirconium oxide, ZrO 2 . Simulated datasets for NAEI-BRA training are generated using the numerical Adams-Bashforth method, ensuring robust numerical correctness and comprehensive coverage of the controlling parameter ranges. The simulation results of the proposed NAEI-BRA closely align with the numerical findings across numerous test scenarios, exhibiting minimal errors and significant consistency. Furthermore, the precision and reliability of the developed NAEI-BRA are meticulously verified through the analysis of mean square error, error histograms, autocorrelation analysis, and regression analysis, yielding results for the BTHN-DFD model that provide compelling evidence of the proposed technique's predictive reliability and efficacy.