用于求解非线性方程连续变形算法的D2*三角剖分

The $D_2^* $-Triangulation for Continuous Deformation Algorithms to Compute Solutions of Nonlinear Equations

SIAM Journal on Optimization · 1993
被引 1
ABS 3

中文导读

提出一种新的D2*三角剖分,用于连续变形算法求解非线性方程,可任意选择正偶数为网格细化因子,在单纯形数量上优于K2*和J2*剖分,数值测试表明算法效率更高。

Abstract

A new triangulation of continuous refinement of grid size of $( 0,1 ] \times R^n $ for use in a continuous deformation algorithm to compute solutions of nonlinear equations is proposed. It is called the $D_2^* $-triangulation. One can choose any positive even integer as a factor of refinement of grid size of this triangulation. The author proves that the $D_2^* $-triangulation is superior to the $K_2^* $-triangulation and $J_2^* $-triangulation in the number of simplices. Numerical tests show that the continuous deformation algorithm based on the $D_2^* $-triangulation is indeed much more efficient.

非线性方程连续变形算法三角剖分数值计算