Behavioral modeling of microelectromechanical systems (MEMS) with statistical performance-variability reduction and sensitivity analysis

An approach to behavioral modeling of microelectromechanical systems (MEMS) is presented emphasizing robust design that minimizes the effects of device parametric variability on overall performance. Using a novel application of Taguchi experimental design and statistical process-control methods, statistical performance-variability reduction and parametric sensitivity analysis are studied. Taguchi performance-variability reduction and parametric sensitivity analyses introduce design-for-manufacturing into behavioral modeling. An example is given for the robust behavioral modeling of a microelectromechanical laterally driven electrostatic-comb microresonator. Applications of the behavioral-modeling approach to high-performance design, manufacturing yield optimization, and operational reliability assessment are also given.