Computational intelligence approach using ANN and FPSO for heat transfer prediction in ternary hybrid nanofluid flow over a stretching sheet
Hemant Kumar, Sawan Kumar Rawat, Moh Yaseen, Manish PantPurpose
The novelty of this research is to investigate the heat transfer characteristics of ternary hybrid nanofluid (NF) flow within a porous medium affected by a magnetic field over a stretching sheet. The distinctive aspect of the work lies in examining the combined effects of thermal radiation, the Cattaneo–Christov heat flux model, suction/injection and the Biot number on the transport phenomena of ternary hybrid NFs.
Design/methodology/approach
The appropriate similarity transformations are applied to reduce the governing partial differential equations (PDEs) into a system of ordinary differential equations. The study further explores the applicability of soft computing approaches in predicting the complex behavior of the system influenced by multiple interacting parameters, including magnetic field, surface suction/injection, thermal radiation, Biot number, nonuniform heat source/sink and porous medium effects. To handle the resultant nonlinear system numerically, the “bvp4c function” in MATLAB is used. After that, an artificial neural network (ANN) and a fuzzy particle swarm optimization (FPSO) technique are used to estimate the Nusselt number values at the stretched sheet with accuracy.
Findings
Furthermore, it is seen that when the radiation parameter (Rd) varies from 0.1 to 2, the heat transfer rate rises by 203.84% and 226.76% for suction parameters S = 1.1 and S = 1.5, respectively. The conclusion highlights the potential of ANN and FPSO algorithms in heat transfer assessments by showing that they provide efficient solutions for physical issues.
Originality/value
The present research extends its scope to examine the effectiveness of soft computing techniques in analyzing fluid flow models that involve heat transfer phenomena.