Mixed bioconvection of nanofluid of oxytactic bacteria through a porous cavity with inlet and outlet under periodic magnetic field using artificial based on

Abstract

The comprehension of microorganisms' responses to magnetic field and fluid motion offers potential for the advancement of targeted drug delivery systems or medical interventions reliant on biological fluids. In this paper, the mixed bioconvection flow of nanofluid of oxytactic bacteria has been investigated through a porous cavity with inlet and outlet ports under the impact of periodic magnetic field. All the walls of the cavity are fixed at the constant high temperature. The proposed problem has been modeled first and then simulated using the finite element method. The computed computational fluid dynamics results have been analyzed for the several important controlling parameters. It is observed that there is at least 65% increment on heat transfer between the lowest and highest Peclet numbers. Further, regression analysis was conducted using the LightGBM algorithm. Careful parameter tuning was performed to avoid overfitting, ensuring that the model did not memorize the training data. According to the R 2 values used for regression analysis, in the 18 datasets used for the performance metric comparison of the artificial intelligence model, a total of 18 targets were tried to be predicted for different Richardson number (0.1, 1, and 10) and Lewis Number (0.1, 1, and 10) values for each motile microorganisms, temperature and oxygen concentration values specified in the datasets, and a minimum accuracy of 95% and a maximum accuracy of 98% were obtained. These findings demonstrate the capability of the LightGBM algorithm to accurately predict the target variable within a high range of accuracy for the given datasets.