Scholarlink

<p>Artificial Intelligence (AI) significantly advances fluid dynamics by accelerating computing procedures and enhancing accuracy while reducing reliance on empirical experimentation. Scientific progress is accelerated by its ability to enable researchers to predict complex fluid phenomena and improve designs. In a present investigation, we employed supervised AI-based nonlinear autoregressive exogenous (NARX) utilized with combination of Levenberg–Marquardt algorithm (LMA) is frequently referred to (NARX-LMA) to examine the mono and hybrid nanofluid over shrinking surface with thermal radiation (MHNFTR). This paper addresses the investigation of magnetohydrodynamics (MHD) incompressible flow of fluid including hydro nanoparticles forming the colloidal combination with base fluid. The governing system of partial differential equations (PDEs) for MHNFTR is transformed into set of ordinary differential equations (ODEs) by employing appropriate transformation. The Adam numerical approach generates datasets for MHNFTR varying parameters Prandtl number (Pr), radiation parameter (<i>R</i>), magnetic parameter (<i>M</i>), and <i>A</i>1, <i>A</i>2, and <i>A</i>3 for velocity and temperature fields. These datasets are further used by the NARX-LMA to estimate numerical results for MHNFTR in seven scenarios. The AI-based approach displays high efficiency, and the outcomes show a strong correlation with the reference data. Error histograms, mean square error (MSE), and absolute error are used to validate performance, and MSE of between E-4 and E-9 is achieved.</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 investigation with novel multi-layer neural networks for mono-hybrid nanofluid model conveying appliance of solar energy

  • 期刊论文

  • 发表日期：

    2025年7月5日

查看全文
  • Zahoor Shah,
  • Maryam Jawaid,
  • Waqar Azeem Khan,
  • Mehboob Ali,
  • Taseer Muhammad

摘要

Artificial Intelligence (AI) significantly advances fluid dynamics by accelerating computing procedures and enhancing accuracy while reducing reliance on empirical experimentation. Scientific progress is accelerated by its ability to enable researchers to predict complex fluid phenomena and improve designs. In a present investigation, we employed supervised AI-based nonlinear autoregressive exogenous (NARX) utilized with combination of Levenberg–Marquardt algorithm (LMA) is frequently referred to (NARX-LMA) to examine the mono and hybrid nanofluid over shrinking surface with thermal radiation (MHNFTR). This paper addresses the investigation of magnetohydrodynamics (MHD) incompressible flow of fluid including hydro nanoparticles forming the colloidal combination with base fluid. The governing system of partial differential equations (PDEs) for MHNFTR is transformed into set of ordinary differential equations (ODEs) by employing appropriate transformation. The Adam numerical approach generates datasets for MHNFTR varying parameters Prandtl number (Pr), radiation parameter (R), magnetic parameter (M), and A1, A2, and A3 for velocity and temperature fields. These datasets are further used by the NARX-LMA to estimate numerical results for MHNFTR in seven scenarios. The AI-based approach displays high efficiency, and the outcomes show a strong correlation with the reference data. Error histograms, mean square error (MSE), and absolute error are used to validate performance, and MSE of between E-4 and E-9 is achieved.