Analysis on Vibration Performance of Slit-Resonant Beam Based on BP Neural Network

Influence of different slit sizes on performance like the natural frequency of slit-resonant beam in a MEMS resonant pressure sensor were investigated. The neural network model is established and trained by sufficient data obtained by Finite Element Method, and is used to obtain effects of different slit sizes on the performance of slit-resonant beam, which verified the results obtained by the theoretical vibration model of double clamped resonant beam with slit structure. The results reveal the different influences of slit size, which can be used as reference for design and optimization of the slit structure and sensor performances.