arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2303.01725math.NAcs.NAmath.AP
收藏

时间分数多孔介质方程:Erdélyi-Kober积分方程,紧支撑解和数值方法

Time-fractional porous medium equation: Erdélyi-Kober integral equations, compactly supported solutions, and numerical methods

Belen López, Hanna Okrasińska-Płociniczak, Łukasz Płociniczak, Juan Rocha

摘要:

时间分数阶多孔介质方程是许多水文、物理和化学流动的重要模型。我们研究了它的自相似解,它构成了许多重要的实验测量情况的轮廓。证明了在一维情形下一般初边值问题存在唯一解。当补充边界条件的物理模型,该问题表现出自相似的解决方案描述的Erdélyi-Kober分数算子的使用。使用向后打靶法,我们表明,存在一个唯一的解决方案,我们的问题。打靶法不仅在推导理论结果时有用。我们利用它来设计一个有效的数值方案来解决管理问题,以及两种方法离散的Erdélyi-Kober分数阶导数。由于后者是一个非局部算子,它的数值实现必须包括一些截断。我们找到正确的截断制度,并证明了几个误差估计。此外,还可以用后向打靶法来求解主问题,并给出了收敛性证明。主要困难在于扩散系数的简并性。我们通过一些正则化来克服它。我们的研究结果补充了数值模拟,验证了理论研究结果。

英文摘要:

The time-fractional porous medium equation is an important model of many hydrological, physical, and chemical flows. We study its self-similar solutions, which make up the profiles of many important experimentally measured situations. We prove that there is a unique solution to the general initial-boundary value problem in the one-dimensional setting. When supplemented with boundary conditions from the physical models, the problem exhibits a self-similar solution described with the use of the Erdélyi-Kober fractional operator. Using a backward shooting method, we show that there exists a unique solution to our problem. The shooting method is not only useful in deriving the theoretical results. We utilize it to devise an efficient numerical scheme to solve the governing problem along with two ways of discretizing the Erdélyi-Kober fractional derivative. Since the latter is a nonlocal operator, its numerical realization has to include some truncation. We find the correct truncation regime and prove several error estimates. Furthermore, the backward shooting method can be used to solve the main problem, and we provide a convergence proof. The main difficulty lies in the degeneracy of the diffusivity. We overcome it with some regularization. Our findings are supplemented with numerical simulations that verify the theoretical findings.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑