Research work · 2013
System‐Level Modeling of MEMS by Means of Model Order Reduction (Mathematical Approximations) – Mathematical Background
Lihong Feng, Peter Benner, Jan Gerrit Korvink
Advanced micro & nanosystems
Abstract
This chapter considers system-level modeling of microelectromechanical systems (MEMS), where model order reduction (MOR) methods based on moment matching play an important role. Furthermore, it also introduces the basic technique of the Gramian-based MOR methods. It first introduces the basic mathematical definitions and concepts, which are necessary to understand all the model reduction methods. Some numerical algorithms are the core of the moment-matching MOR methods, which are explained in the chapter for a better and easier application of moment-matching MOR. The chapter introduces some knowledge on system theory to understand the Gramian-based MOR methods, presents the basic idea of MOR, and explains the motivation and implementation for each kind of MOR method. It then analyzes the error estimation, stability, and passivity properties of the reduced models, and discusses an issue about nonzero initial condition. The chapter finally briefly mentions MOR for more complex systems. Controlled Vocabulary Terms micromechanical devices; numerical analysis; power system simulation; system theory