基于多层简易矩阵稀疏算法的复杂物体建模优化的研究  

Research on complex object modeling optimization based on multilayer simple sparse matrix algorithm

作　　者：杨全海[1] 

机构地区：[1]陕西职业技术学院,陕西西安710000

出　　处：《现代电子技术》2015年第22期14-16,共3页Modern Electronics Technique

摘　　要：为了提高三维大尺寸复杂物体电磁建模的精确求解的效率,改进了自适应交叉算法(ACA)基础上的多层简易矩阵稀疏算法(MLSSM),通过对算法的理论及实现过程分析,并在实验验证中表明:应用的改进ACA算法计算效率比矩量法逐点计算显著提高;改进的MLSSM的内存需求减少了1 2左右,矩阵构造过程速度有了明显提高;改进的MLSSM降低了计算复杂度,迭代求解过程速度有了明显提高。算法对比结果看出改进的MLSSM在降低计算复杂度方面占据的优势,在分析半空间上大目标复杂物体优势明显。这一研究对于复杂物体的电磁建模的优化有一定的理论和应用意义。The exact solution efficiency of large-size three-dimensional electromagnetic modeling of complex objects was im- proved. The improved mode of MLSSM （muhilayer simple sparse matrix algorithm） based on ACA （adaptive crossover algo- rithm） was adopted. By the theory and realization process analysis, the algorithm was verified in the experiment. The experimen- tal results show that the improved ACA algorithm can improve computational efficiency more significantly than point-by-point computation of the moment method, the memory requirements of improved MLSSM is reduced by about half, and the matrix structure is improved significantly. The speed of iterative solution process is greatly increased. The comparative results indicate that the MLSSM has the advantages of reducing computational complexity and analyzing large complex objects in semi-space. The study has certain significance in theory and application for the optimization of electromagnetic modeling of complex objects.

关 键 词：电磁建模 精确求解 复杂物体建模 多层简易矩阵稀疏算法 自适应交叉算法 

分 类 号：TN911-34[电子电信—通信与信息系统] TN911.7[电子电信—信息与通信工程]