【作者】 赵延文； 张雪峰； 陆田； 杨颖怡；

【Author】 Zhao Yan-Wen1 Zhang Xue-Feng1,2 Lu Tian1 Yang Ying-Yi1 (1. School of Electronic Engineering, University of Electronic Science and Technology of China, Chengdu Sichuan 610054, China. 2. College of Material Engineering, University of Pan Zhihua, Pan Zhihua Sichua, 617000.)

【机构】 电子科技大学电子工程学院； 攀枝花学院材料工程学院；

【摘要】 矩阵预条件技术广泛用于频域积分方程矩量法、有限元法等,以提高对应矩阵方程迭代解法的收敛速度。本文将近场预条件技术和不完全LU分解(ILU)预条件技术用于时域积分方程时间步进算法。通过计算实例表明,将包括近场预条件技术和不完全LU分解(ILU)预条件技术在内的矩阵预条件技术用于时域积分方程时间步进(MOT)算法时,具有简单易行、应用灵活、高效率、低内存需求的特点。更多还原

【Abstract】 Preconditioners have been developed to accelerate the convergence rate of iterative approachs for solving matrix equations, arising from the method of moments (MoM) of the frequency-domain integral equations and from the finite element methods. In this paper, the near-field preconditioner and incomplete LU (ILU) preconditioner are adopted to dramatically reduce the iteration count of the iterative solvers for solving the sparse matrix equations, arising from the MOT scheme of the time-domain integral equations. Numerical results demonstrate that these preconditoners have low memory and CPU requirements, and are easily implemented.更多还原