Parametric model of resonant differential MEMS pressure sensors: enabling rapid structural optimization and automated layout design

Abstract This paper aims to enhance the efficiency of sensor structural optimization design and achieve rapid performance evaluation and automatic layout drawing. To this end, we propose a parameterized model of a resonant differential MEMS pressure sensor (PRDMP). It can complete sensitivity calculations within 1 ms, far faster than traditional finite element analysis (FEA) methods. Additionally, compared to FEA results, its accuracy exceeds 90 <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" overflow="scroll"> <mml:mrow> <mml:mi mathvariant="normal">%</mml:mi> </mml:mrow> </mml:math> . Furthermore, we reconstructed the uncertainty analysis part of PRDMP based on an error model. Compared to the Monte Carlo method used in existing studies, our method is faster and yields more stable results. Benefiting from the speed and accuracy of PRDMP, we achieved, for the first time, the multi-parameter collaborative automatic optimization of this sensor. Optimization results show that sensitivity increased by 36.2 <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" overflow="scroll"> <mml:mrow> <mml:mi mathvariant="normal">%</mml:mi> </mml:mrow> </mml:math> while uncertainty decreased by 15.8 <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" overflow="scroll"> <mml:mrow> <mml:mi mathvariant="normal">%</mml:mi> </mml:mrow> </mml:math> . Finally, we developed a sensor performance analysis and automatic layout drawing tool based on PRDMP, further enhancing design efficiency.