Computational Challenges in Geometrical Flows: Numerical Methods and Analysis, Algorithmic Development and Software Engineering

几何流中的计算挑战：数值方法和分析、算法开发和软件工程

• 批准号: 0410266
• 负责人: Xiaobing Feng
• 金额: --
• 依托单位: University of Tennessee Knoxville
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2004
• 资助国家: 美国
• 起止时间: 2004-08-01 至 2008-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0410266&HistoricalAwards=false
• 关键词: Computational; Challenges; Geometrical; Flows; Numerical

This research is an attempt to conduct an extensive and comprehensive numerical study of geometrical flows, such as the mean curvature flow, the inverse mean curvature flow, the Gauss curvature flow, the surface diffusion flow, the Willmore flow, and the Ricci flow, from differential geometry, fluid mechanics, materials science, and cosmology, and to address theoretically and numerically challenging issues arising from geometrical flow computations. The goal of this project is to develop accurate, robust, and efficient adaptive finite element discretization methods, parallel iterative solution algorithms and computer codes for computing geometrical flows based on both level set and phase field formulations. The investigator aims to carry out a balanced numerical study for the geometrical flows by emphasizing both qualitative analysis and quantitative computation, and thus to provide reliable computational tools for discovering and analyzing fine properties such as dynamics of the singularities of the geometrical flows, which often are difficult and even may not be possible to predicate and characterize by analytical means. The methods and algorithms resulting from this research will have the following attractive features: high accuracy, strong stability, low cost, and high efficiency. In addition, the proposed methods are also capable of accurately and efficiently approximating the geometrical flows not only before but also beyond the onset of singularities.As critical applications from fluid mechanics, cosmology, and materials science are directly tied to the solutions of geometrical flows, it is expected that successful completion of the proposed research has the potential to significantly impact these applied sciences not only by presenting new methods for solving underlying mathematical problems but also providing insights for the understanding of each of these applications. Furthermore, the methods to be developed will find applications in other fields such as cell biology, geophysics, image processing, and computer vision. The educational component of the project consists of graduate graduate course development, training and mentoring both graduate and undergraduate students through the project.

本研究试图从微分几何、流体力学、材料科学和宇宙学的角度对几何流动进行广泛而全面的数值研究，包括平均曲率流、逆平均曲率流、高斯曲率流、表面扩散流、Willmore流和Ricci流，并解决几何流动计算所产生的理论和数值挑战问题。该项目的目标是开发准确、稳健和高效的自适应有限元离散方法、并行迭代求解算法和计算机代码，用于计算基于水平集和相场公式的几何流动。研究者的目标是通过强调定性分析和定量计算，对几何流动进行平衡的数值研究，从而为发现和分析几何流动奇异性的动力学等精细性质提供可靠的计算工具，这些性质往往很难甚至可能无法用分析手段来预测和表征。本研究提出的方法和算法具有精度高、稳定性强、成本低、效率高等特点。由于流体力学、宇宙学和材料科学的关键应用都直接与几何流的解联系在一起，因此，拟议的研究的成功完成有望对这些应用科学产生重大影响，不仅为解决基本的数学问题提供了新的方法，而且还为理解这些应用提供了见解。此外，即将开发的方法还将在其他领域得到应用，如细胞生物学、地球物理学、图像处理和计算机视觉。该项目的教育部分包括通过该项目开发研究生课程、培训和指导研究生和本科生。