Gaussian Regression Process Modeling of Bioconvective Nanofluid Flow in a Magnetically Driven Square Cavity
4th Cognitive Models and Artificial Intelligence Conference, AICCONF 2026, Prague, Çek Cumhuriyeti, 24 - 25 Nisan 2026, (Tam Metin Bildiri)
- Yayın Türü: Bildiri / Tam Metin Bildiri
- Doi Numarası: 10.1109/aicconf69182.2026.11600725
- Basıldığı Şehir: Prague
- Basıldığı Ülke: Çek Cumhuriyeti
- Anahtar Kelimeler: Bioconvection, Gaussian regression process modeling, induced magnetic field, nanofluid
- TED Üniversitesi Adresli: Evet
Özet
This study develops a Gaussian process regression (GPR) model to predict the buoyancy-driven bioconvection behavior of an Fe3O4-water nanofluid containing magnetotactic bacteria in a unit square cavity subject to externally applied and induced magnetic fields. The steady, two-dimensional governing equations are formulated in dimensionless stream function-vorticity form, coupled with equations for temperature, iron concentration, bacteria density, and magnetic potential. A radial basis function (RBF) collocation method is employed to generate a comprehensive dataset covering wide ranges of Rayleigh, bioconvective Rayleigh, Peclet, Lewis, Hartmann, magnetic Reynolds, and buoyancy ratio numbers. For each parameter combination, the average Nusselt number, Sherwood number and bacteria density along the heated wall are recorded. Using this dataset, data-driven GPR model is constructed for each of the output quantities. Model performance is assessed through mean squared error (MSE) metrics, hyperparameter sensitivity analysis, and residual plots. The results demonstrate that the proposed GPR model with RBF kernel accurately captures the multiphysics relationships, achieving test MSEs on the order of 10-3 for the averages. The developed framework offers an alternative prediction modeling for bioconvection systems.