| تعداد نشریات | 45 |
| تعداد شمارهها | 1,562 |
| تعداد مقالات | 19,119 |
| تعداد مشاهده مقاله | 66,172,319 |
| تعداد دریافت فایل اصل مقاله | 24,041,699 |
A Regularized Expansion-Iterative Method for Stable Solutions of Ill-Posed Integral Equations of the First Kind | ||
| Computational Methods for Differential Equations | ||
| مقالات آماده انتشار، پذیرفته شده، انتشار آنلاین از تاریخ 28 آبان 1405 اصل مقاله (1.1 M) | ||
| نوع مقاله: Research Paper | ||
| شناسه دیجیتال (DOI): 10.22034/cmde.2026.72372.3654 | ||
| نویسندگان | ||
| Saeed Hatamzadeh1؛ Tofigh Allahviranloo2؛ Zahra Masouri* 3 | ||
| 1Department of Electrical and Electronics Engineering, Faculty of Engineering and Natural Sciences, Istinye University, Istanbul, Turkey. | ||
| 2Research Center of Performance and Productivity Analysis, Istinye University, Istanbul, Turkey. | ||
| 3Department of Mathematics, Isl.C., Islamic Azad University, Islamshahr, Iran. | ||
| چکیده | ||
| This paper presents a Regularized Expansion–Iterative (REI) method for obtaining stable numerical solutions of Fredholm integral equations of the first kind, which are archetypal examples of ill-posed inverse problems. The proposed approach combines a block-pulse expansion with an iterative refinement process that incorporates an intrinsic Tikhonov-type regularization and a relaxation control parameter. Unlike conventional spectral or variational regularization schemes, the REI algorithm avoids explicit matrix inversion and singular-value decomposition, operating instead through simple matrix–vector updates with built-in stabilization at each iteration. Theoretical analysis establishes that the iteration sequence remains bounded, exhibits monotone error reduction, and converges to a regularized solution equivalent to the classical Tikhonov limit under appropriate scaling of regularization parameter λ. Despite this theoretical equivalence, the REI framework offers significant computational advantages: it eliminates the need to form or invert normal equations, ensures automatic regularization through iteration, and provides faster and smoother convergence with reduced computational cost. Numerical experiments on four canonical benchmark problems with Gaussian, rational, absolute-value, and exponential kernels demonstrate that REI achieves accurate and stable reconstructions with reasonable errors, even under moderate data noise. Parameter-sweep analyses confirm the predictable influence of λ, relaxation parameter ω, and noise level η, and show that the method remains numerically stable for all tested configurations. Owing to its simplicity, robustness, and independence from spectral decomposition, the REI framework provides a practical and efficient alternative to classical regularization schemes for large-scale ill-posed integral equations. | ||
| کلیدواژهها | ||
| Ill-posed Fredholm integral equations of the first kind؛ Expansion-Iterative regularization؛ Relaxation parameter؛ Error and convergence analysis؛ Stability and noise robustness | ||
|
آمار تعداد مشاهده مقاله: 8 تعداد دریافت فایل اصل مقاله: 6 |
||