An Extrapolation Accelerated Multiscale Newton-MG Method for Fourth-Order Compact Discretizations of Semilinear Poisson Equations

Hongling Hu,Ming Li,Kejia Pan,Pinxia Wu
DOI: https://doi.org/10.1016/j.camwa.2022.03.003
IF: 3.218
2022-01-01
Computers & Mathematics with Applications
Abstract:An extrapolation accelerated multiscale Newton-multigrid (EMNMG) method is proposed to solve two-dimensional semilinear Poisson equations. The nine-point fourth-order compact schemes are used to approximate the nonlinear Poisson equations. In order to accelerate Newton-MG method for calculating the finite difference (FD) solution on the finest grid, a quite good initial guess is constructed from the fourth-order FD solutions at two coarse levels by using Richardson extrapolation and bi-quartic polynomial interpolation, which greatly reduces the number of Newton iterations required. A completed extrapolation technique is adopted to generate a sixth-order extrapolated solution on entire finest grid cheaply. Numerical results are given to show that our proposed EMNMG algorithm can achieve high accuracy and keep less cost simultaneously.
What problem does this paper attempt to address?