آمار
| تعداد نشریات | 45 |
| تعداد شمارهها | 1,531 |
| تعداد مقالات | 18,683 |
| تعداد مشاهده مقاله | 61,463,753 |
| تعداد دریافت فایل اصل مقاله | 22,215,690 |
Enhanced MLS-NN Collocation for Fractional EHD Flows in Cylindrical Conduits | ||
| Computational Methods for Differential Equations | ||
| مقالات آماده انتشار، پذیرفته شده، انتشار آنلاین از تاریخ 10 مرداد 1405 اصل مقاله (2.43 M) | ||
| نوع مقاله: Research Paper | ||
| شناسه دیجیتال (DOI): 10.22034/cmde.2026.69521.3427 | ||
| نویسندگان | ||
| Parisa Rahimkhani* 1؛ Zahra Khoddami Maraghi* 2؛ Ehsan Arshid2 | ||
| 1Faculty of Science, Mahallat Institute of Higher Education, Mahallat, Iran. | ||
| 2Faculty of Engineering, Mahallat Institute of Higher Education, Mahallat, Iran. | ||
| چکیده | ||
| This study presents a hybrid numerical framework that combines the enhanced moving least squares (MLS) approximation, neural networks, and spectral methods to solve a fractional-order electrohydrodynamics flow model in a cylindrical conduit. The proposed method transforms the governing problem into a nonlinear system of algebraic equations. The enhanced MLS technique facilitates the construction of smooth approximations for scattered or irregular data distributions, while the neural network component provides a flexible mechanism for modeling the nonlinear and memory-dependent behavior inherent in fractional-order systems. By embedding MLS-based shape functions as activation mechanisms, the framework effectively captures the solution. Furthermore, a collocation strategy based on the roots of the Legendre polynomials is employed to minimize the residual associated with the underlying fractional operator. This approach improves numerical accuracy and robustness, as verified by the convergence analysis provided in this study. To show the effectiveness of the proposed method, some numerical results are provided. | ||
| کلیدواژهها | ||
| Fractional Electrohydrodynamics flow؛ Enhanced moving least squares approximation؛ Neural networks؛ Spectral method | ||
|
آمار تعداد مشاهده مقاله: 27 تعداد دریافت فایل اصل مقاله: 25 |
||