论文部分内容阅读
Rabbat、Sangiovanni和Hsieh提出了适用于大规模非线性电路时域分析的多级牛顿算法(MLNA)。本文对MLNA提出两个改进:1.用子网络的端点电压作为子网络的激励。2.使用灵敏度网络求取子网络对雅可比矩阵的贡献。从而节省了建立雅可比矩阵的时间。
Rabbat, Sangiovanni and Hsieh proposed a multistage Newton algorithm (MLNA) for time domain analysis of large scale nonlinear circuits. This paper presents two improvements to MLNA: 1. Use the sub-network’s endpoint voltage as the sub-network stimulus. 2. Use the sensitivity network to find the contribution of the sub-network to the Jacobian matrix. Thus saving the time to establish Jacobian matrix.