1. Neural intellectual computing systems for the analysis of thermally stratified mixed convective micropolar liquid with the interaction of thermal diffusive nanofluid over a heated sheet.
- Author
-
Alghamdi, Metib, Zamir, Tayyab, Akbar, Noreen Sher, and Muhammad, Taseer
- Subjects
ARTIFICIAL neural networks ,FLUID flow ,SIMILARITY transformations ,ORDINARY differential equations ,COMPUTER systems ,NANOFLUIDICS - Abstract
This study focuses on investigating entropy generation in a micropolar fluid flow over a transversely heated sheet. The governing equations of the specific model proposed in this article are transformed into a set of ordinary differential equations (ODEs) using similarity transformations, leading to a non-spatial arrangement. A BVP4C approach is then employed to solve this system of non-spatial ODEs. The Levenberg–Marquardt procedure, known for its effectiveness in artificial neural networks, is utilized to obtain a numerical solution for the entropy generation in micropolar liquid (EGML). This solution is achieved through regression plots, state transition measures, histogram representations, and mean squared errors. In this research, an EGML-based fluid flow problem is examined using an innovative application of an intelligent computer system that leverages neural structures and the Levenberg–Marquardt algorithm (NN-LMA). Through the BVP4C approach, data collection is conducted to facilitate the implementation of the NN-LMA. The EGML results obtained for various scenarios are evaluated using the exercise, acceptance, and trial processes of NN-LMA. A comparison is then made between these results and a reference dataset to validate the accuracy and efficiency of the prospective algorithm NN-LMA for investigating fluid problems associated with EGML. The study delves into how various non-dimensional factors influence flow patterns, temperature distributions, concentration profiles, and velocity profiles. State transition dynamics, regression analysis, mean square error calculations, and error histogram investigations are employed to verify the validity of the proposed NN-LMA method for solving the EGML. Overall, the results obtained through these analyses successfully demonstrate the efficacy and accuracy of the NN-LMA approach in addressing EGML-related fluid flow issues. The framework's validity is demonstrated by the strong congruence of recommended results with reference solutions, and accuracy of 1 0 - 11 to 1 0 - 10 is also attained. [ABSTRACT FROM AUTHOR]
- Published
- 2025
- Full Text
- View/download PDF