Scholarlink

<p>The present article contains a neural networking analysis of Powell–Eyring (PE) fluid flow directed toward a flat surface with suction and injection processes. Heat and mass transfer traits are well-thought-out simultaneously with free stream effect. Flow is formulated in terms of partial differential equations, and the theory of Lie symmetry groups is used to construct the particular set of transformations for reduction of order. The reduced equations are transformed into a system of first-order ordinary differential equations as an initial value problem. The shooting methodology is adopted to report the thermal flow field by evaluating the Nusselt number and temperature. The 4 × 90 matrix is constructed for the Prandtl number, PE fluid parameter, suction parameter and temperature power law index, while the 1 × 90 matrix is developed for the Nusselt number. The 75% (67), 10% (9), and 15% (14) are slotted for training, validation, and testing, respectively, and in hidden layer 15 neurons are used. Levenberg–Marquardt backpropagation is used to train the neural networking model. Regression and mean square analysis are carried out to measure the accuracy of the constructed neural network for Nusselt number. Following the predicted values, the Nusselt number admits inciting values toward the temperature power law index, Prandtl number, and suction parameter while contradictory is the case for the PE fluid parameter.</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

Neural networking-based decisions on Powell–Eyring fluid flow with identical stretched porous and free stream conditions

  • 期刊论文

  • 发表日期：

    2025年2月28日

查看全文
  • Khalil Ur Rehman,
  • Wasfi Shatanawi,
  • Lok Yian Yian

摘要

The present article contains a neural networking analysis of Powell–Eyring (PE) fluid flow directed toward a flat surface with suction and injection processes. Heat and mass transfer traits are well-thought-out simultaneously with free stream effect. Flow is formulated in terms of partial differential equations, and the theory of Lie symmetry groups is used to construct the particular set of transformations for reduction of order. The reduced equations are transformed into a system of first-order ordinary differential equations as an initial value problem. The shooting methodology is adopted to report the thermal flow field by evaluating the Nusselt number and temperature. The 4 × 90 matrix is constructed for the Prandtl number, PE fluid parameter, suction parameter and temperature power law index, while the 1 × 90 matrix is developed for the Nusselt number. The 75% (67), 10% (9), and 15% (14) are slotted for training, validation, and testing, respectively, and in hidden layer 15 neurons are used. Levenberg–Marquardt backpropagation is used to train the neural networking model. Regression and mean square analysis are carried out to measure the accuracy of the constructed neural network for Nusselt number. Following the predicted values, the Nusselt number admits inciting values toward the temperature power law index, Prandtl number, and suction parameter while contradictory is the case for the PE fluid parameter.