Suppr超能文献

单位球体内狭窄捕获问题的全局最优体积陷阱排列。

Globally optimal volume-trap arrangements for the narrow-capture problem inside a unit sphere.

机构信息

Department of Mathematics and Statistics, University of Saskatchewan, Saskatoon S7N 5E6, Canada.

出版信息

Phys Rev E. 2019 Jan;99(1-1):012109. doi: 10.1103/PhysRevE.99.012109.

Abstract

The determination of statistical characteristics for particles undergoing Brownian motion in constrained domains has multiple applications in various areas of research. This work presents an attempt to systematically compute globally optimal configurations of traps inside a three-dimensional domain that minimize the average of the mean first passage time (MFPT) for the narrow capture problem, the average time it takes a particle to be captured by any trap. For a given domain, the mean first passage time satisfies a linear Poisson problem with Dirichlet-Neumann boundary conditions. While no closed-form general solution of such problems is known, approximate asymptotic MFPT expressions for small traps in a unit sphere have been found. These solutions explicitly depend on trap parameters, including locations, through a pairwise potential function. After probing the applicability limits of asymptotic formulas through comparisons with numerical and available exact solutions of the narrow capture problem, full three-dimensional global optimization was performed to find optimal trap positions in the unit sphere for 2≤N≤100 identical traps. The interaction energy values and geometrical features of the putative optimal trap arrangements are presented.

摘要

在约束域中进行布朗运动的粒子的统计特性的确定在各个研究领域都有多种应用。这项工作试图系统地计算三维域内陷阱的全局最优配置,以最小化窄捕获问题的平均首次通过时间 (MFPT) 的平均值,即粒子被任何陷阱捕获所需的平均时间。对于给定的域,平均首次通过时间满足具有狄利克雷-诺伊曼边界条件的线性泊松问题。虽然此类问题没有已知的封闭形式的一般解,但已经找到了单位球中小陷阱的近似渐近 MFPT 表达式。这些解通过与窄捕获问题的数值和可用精确解的比较,明确依赖于陷阱参数,包括位置,通过一个成对势函数。在通过与窄捕获问题的数值和可用精确解的比较来探测渐近公式的适用范围限制之后,对单位球中的 2≤N≤100 个相同陷阱进行了完整的三维全局优化,以找到最优陷阱位置。给出了相互作用能值和假定的最优陷阱排列的几何特征。

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验