摘要
提出了一种采用边界检测的自适应区域分解时域有限差分(FDTD)方法:将待解问题分解为若干独立的子区域;在各个子区域之间通过连接边界条件,进行自适应检测;当检测到波传播至检测边界上时,相应子区域被激活,并进行传统的FDTD迭代计算,否则,该子区域将不参与场值迭代.通过这样一种边界检测信息传递方案,把各个子区域联系起来,获得原问题的解.实际仿真结果表明,这种方法在解决复杂问题和加快计算时间方面是有效的.
An extension of the classical boundary detection method, is proposed FDTD, adaptive domain-decomposition FDTD (ADD-FDTD) using a to solve multiple sub-region problems. In ADD-FDTD, the problem domain is divided into several independent FDTD sub-regions. A boundary condition is applied to boundaries between sub-regions with adaptive detection. In each sub-region, FDTD iteration is activated only if the propagated wave reaches its boundary, otherwise no calculation is performed. Simulation results show that ADD-FDTD is efficient in both computational complexity and computational time.
出处
《应用科学学报》
CAS
CSCD
北大核心
2005年第5期493-496,共4页
Journal of Applied Sciences
基金
国家"863"高技术研究发展计划(2002AA123031)
国家杰出青年科学基金(60225001)
海外青年合作研究基金资助项目(60028101)
关键词
自适应边界检测
区域分解
时域有限差分法
adaptive boundary testing
domain decomposition
finite-difference time-domain