论文部分内容阅读
针对含大规模输入输出端口的MEMS结构,提出分块叠加Arnoldi方法提取其宏模型,针对MEMS器件中广泛存在几何形状相同而初始方位不同的结构,采用矩阵坐标变换实现宏模型的角度参数化,解决宏模型的重复提取问题。最后,结合一种微型可编程光栅的设计验证了宏模型的效率与精度,与有限元仿真结果进行比较,基于宏模型的MEMS器件系统级仿真结果相对误差小于2%,时域仿真速度提高了45倍。
Aiming at the MEMS structure with large-scale input and output ports, a block superimposed Arnoldi method is proposed to extract its macromodels. For the MEMS devices, the structures with the same geometry and different initial azimuths are widely existed, and the angular parameterization of the macro model is realized by matrix coordinate transformation. Solve the problem of repeated extraction of macro model. Finally, the efficiency and accuracy of the macro model are verified by the design of a micro-programmable grating. Compared with the finite element simulation results, the relative error of system-level simulation results based on the macro-model is less than 2% and the time-domain simulation speed is improved 45 times.