arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2106.07609math.NAcs.NA
收藏

波模型的精确谱导数离散化Finite Difference(ESDDFD)方法:通过自然分数/分数导数表示的通用波视图(或指数波宇宙的视图透镜商店)

The Exact Spectral Derivative Discretization Finite Difference (ESDDFD) Method for Wave Models: A Universal Wave View Through Natural Fractional / Fractal Derivative Representations (or View Lens Shops for The Exponential Wave Universe)

D. P. Clemence-Mkhope

摘要:

提出了一种宇宙的波动观,其中每一种自然现象都配备了自己独特的自然观察透镜。提出了一种基于傅里叶-拉普拉斯变换空间的对流-扩散-反应方程(ADRE)的自相似建模原理及其系统应用,作为一种新的、通用的离散建模范式,用于求解非整数导数、时间尺度和海浪谱特征的ADRE。它的实现是一种新的精确光谱导数离散化有限差分法(ESDDFD),一种通过从ADRE模型获得离散波模型来制作观波透镜的方法。用于构建这些透镜的模板以从FL变换空间中的波签名概率分布函数及其谐波振荡获得的自然导数表示的形式出现;在波动方程的离散数值建模中使用ESDFD方法不需要任何数学导数的先验理论。这个观点的一个主要的数学后果是,所有的概念的数学整数或非整数衍生物有代表性的限制,这样的自然衍生物表示;这和其他后果进行了讨论,并提出了一个简单的整数衍生物扩散反应方程的离散化来说明该方法。由此产生的视图透镜,在ESDFD模型的形式,工作以及在检测本地和非本地德拜或Kohlrausch-Williams-Watts指数模式,只有布朗运动和子扩散在本文中进行了讨论。

英文摘要:

A wave view of the universe is proposed in which each natural phenomenon is equipped with its own unique natural viewing lens. A self-sameness modeling principle and its systematic application in Fourier-Laplace transform space is proposed as a novel, universal discrete modeling paradigm for advection-diffusion-reaction equations (ADREs) across non-integer derivatives, time scales, and wave spectral signatures. Its implementation is a novel exact spectral derivative discretization finite difference method (ESDDFD), a way for crafting wave viewing lenses by obtaining discrete wave models from ADRE models. The template for building these lenses come in the form of natural derivative representations obtained from the wave signature probability distribution function and its harmonic oscillation in FL transform space; use of the ESDDFD method in the discrete numerical modeling of wave equations requires no a-priori theory of any mathematical derivative. A major mathematical consequence of this viewpoint is that all notions of the mathematical integer or non-integer derivatives have representation as limits of such natural derivative representations; this and other consequences are discussed and a discretization of a simple integer derivative diffusion-reaction equation is presented to illustrate the method. The resulting view lenses, in the form of ESDDFD models, work well in detecting both local and non-local Debye or Kohlrausch-Williams-Watts exponential patterns; only Brownian motion and sub-diffusion are discussed in the present article.

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。

↑