Scholarlink

<p>The complicated behaviour of micropolar fluids under the influence of thermophoresis, Brownian motion, and bioconvection generated by gyrotactic microorganisms is thoroughly investigated in this work. We use an innovative dual-methodology framework that blends cutting-edge machine learning methods with conventional numerical analysis. We computationally examine magnetohydrodynamic flow across stretched surfaces buried in porous media, accounting for thermal radiation and viscous heating effects, using MATLAB’s (bvp4c). Our findings clarify the complex connections between important system parameters and their impacts on fluid flow properties, such as buoyancy ratio, thermophoresis parameter, Peclet number, and bioconvection Lewis and Rayleigh numbers. Five essential fluid dynamic parameters are the skin friction coefficient, surface couple stress, Nusselt number, Sherwood number, and motile microorganism transfer rate. We create and validate several linear regression models to improve forecasting capabilities. With mean relative errors of roughly 2.1% and <i>R</i><sup>2</sup> values ranging from 0.816 to 0.825, the regression models show strong predictive power, capturing intricate nonlinear interactions that are usually missed by traditional analytical techniques.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Machine learning-based analysis of bioconvective micropolar fluid flow over a stretching surface with Brownian motion and thermophoresis effects

  • 期刊论文

  • 发表日期：

    2025年11月23日

查看全文
  • Anurag Bhatnagar,
  • Manish Tanwar,
  • Ruchika Mehta

摘要

The complicated behaviour of micropolar fluids under the influence of thermophoresis, Brownian motion, and bioconvection generated by gyrotactic microorganisms is thoroughly investigated in this work. We use an innovative dual-methodology framework that blends cutting-edge machine learning methods with conventional numerical analysis. We computationally examine magnetohydrodynamic flow across stretched surfaces buried in porous media, accounting for thermal radiation and viscous heating effects, using MATLAB’s (bvp4c). Our findings clarify the complex connections between important system parameters and their impacts on fluid flow properties, such as buoyancy ratio, thermophoresis parameter, Peclet number, and bioconvection Lewis and Rayleigh numbers. Five essential fluid dynamic parameters are the skin friction coefficient, surface couple stress, Nusselt number, Sherwood number, and motile microorganism transfer rate. We create and validate several linear regression models to improve forecasting capabilities. With mean relative errors of roughly 2.1% and R2 values ranging from 0.816 to 0.825, the regression models show strong predictive power, capturing intricate nonlinear interactions that are usually missed by traditional analytical techniques.