The accurate and efficient evaluation of Newtonian potentials over general 2-D domains is a subject of great importance for the numerical solution of Poisson's equation and volume integral equations. Complicated domains typically require discretization by unstructured meshes, over which the direct evaluation of the potential by quadrature becomes costly. In this paper, we present a simple and efficient high-order algorithm for computing Newtonian potentials, based on the use of Green's third identity for transforming the volume integral into a collection of boundary integrals, which can be easily handled by the Helsing-Ojala method. As a result, the time cost of the classically expensive near field and self-interaction computations over an unstructured mesh becomes roughly the same as the time cost of the FMM-based far field interaction computation. One of the key components of our algorithm is the high-order 2-D monomial approximation of a function over a mesh element, which is often regarded as an ill-conditioned problem, since it involves the solution of a Vandermonde linear system. In fact, it has long been observed that, when the function is sufficiently smooth, and when a backward stable linear system solver is used, the resulting monomial expansion can approximate the function uniformly to high accuracy. We rigorously formalize this observation in this paper. The performance of our algorithm is illustrated through several numerical experiments.
翻译:对牛顿在一般二维域上的潜力进行准确而有效的评估,对于Poisson的方程式和体积整体方程式的数值解决方案来说,是一个非常重要的主题。复杂域通常需要由非结构化的网间隔板进行分解,而对于这种分层的可能性进行直接评估则成本高昂。在本文中,我们提出了一个简单而高效的牛顿方程式潜力计算高端算法,其依据是使用格林的第三个身份将体积组合转换成一组边界内构件的集合,这很容易由Helsing-Ojala方法处理。因此,典型昂贵的近野外和自对非结构化网间网间计算的时间成本与基于FMM的远地方互动计算的时间成本大致相同。我们算的关键组成部分之一是计算牛顿方程式的功能高端 2-D 单项近似,它通常被视为一个有缺陷的问题,因为它涉及Vandermonde线性系统的解决办法。事实上,我们长期观察到了在非结构化网际轨迹上进行自我调节的精确度观测,而我们又使用了一种稳定的平坦态的平坦度的轨道。