摘要
采用压缩式双共轭梯度算法分析大规模电源/地线网络.首先以稀疏存储结构对大规模的系数矩阵进行压缩处理,然后采用双共轭梯度算法对网络进行模拟.双共轭梯度算法采用2组共轭向量组作为搜索方向,收敛速度快.实验数据表明:在保证精度的情况下,该算法在加快电路网络分析求解效率的同时,大幅度地节省了计算所占用的内存,它适用于分析超大规模的电源/地线网络.
An improved method is proposed based on compressed bi-conjugate gradient (BCG) algorithm to perform efficient static and transient simulations for large-scale power and ground networks circuits. A more efficient compressed storage strategy is used to save the memory, and BCG algorithm is used to analysis the large-scale power and ground networks. Extensive experimental results show that the compressed BCG algorithm gain the significant memory and run-time advantages over the traditional approach and has more powerful capability to deal with the increasing size of power and ground networks.
出处
《计算机辅助设计与图形学学报》
EI
CSCD
北大核心
2007年第10期1259-1262,共4页
Journal of Computer-Aided Design & Computer Graphics
基金
西安-应用材料创新基金(XA-AM-200605)
关键词
电路模拟
电源/地线网络
双共轭梯度算法
压缩
circuit simulations
power and ground networks
bi-conjugate gradient algorithm
compression