Research On The MEMS Bionic Hair Sensor Based On Mode Localization Of The Weakly Coupled Resonators

The micro-electro-mechanical system(MEMS)bionic hair sensor inspired by the insect hair perception system is a typical multifunctional and miniaturized novel sensor,which can detect airflow velocity,acceleration,sound,vibration and other signals,and is expected to be widely used in the airflow perception of micro autonomous systems,motion control,environmental recognition and other intelligent sensing fields.Different from the capacitive,piezoelectric,or piezoresistive sensing principles,this thesis applies the weakly coupled resonators(WCRs)based on the mode localization phenomenon to the structure design of the bionic hair sensor for the first time.Theoretically,compared with the eigenvalue-modulated resonators,the eigenstate-modulated weakly coupled resonators has higher relative mechanical sensitivity and better common mode rejection to ambient fluctuations(such as temperature,pressure,etc.).Therefore,the weakly coupled resonators can effectively improve the comprehensive performance of the bionic hair sensor.This paper mainly studies the signal sensing mechanism of the weakly coupled resonant MEMS)bionic hair sensor,especially the eigenstate modulation mechanism of the weakly coupled resonators.On this basis,the sensitive structrue and the measurement and control circuit system of the bionic hair sensor are designed and optimized.The main research work of the thesis can be summarized as follows:The theories and working principle of the weakly coupled resonant MEMS bionic hair sensor is studied.The mechanical model of the hair structure is established and the sensing mechanism of the cylindrical hair structure to steady-state airflow signals and oscillating airflow signals is theoretically analyzed.The simplified “mass-spring-damping” model of the two-degree-of-freedom(2-Do F)weakly coupled resonant system is established,the output characteristics of the weakly coupled resonant system under three conditions are analyzed.The relative sensitivities of the eigenvalues and eigenstates of the weakly coupled resonant system under two stiffness perturbation scenarios are analyzed and compared.The mode aliasing of the weakly coupled resonant system is studied and the minimum coupling stiffness coefficient is determined.The output linearity and linear range of the 2-Do F weakly coupled resonant system are studied.Aiming at the strong nonlinearity of the weakly coupled resonant system near the zero point,three full-scale linear sensing output metrics are proposed,namely The combination of amplitude difference and amplitude ratios(CADAR),the summation of amplitude ratio(SAR)and the subtraction of reciprocal amplitude ratios(SRAR).The damping characteristics and non-ideal error characteristics of the three full-scale linear sensing output metrics are compared.Numerical calculation and analysis results demonstrate that the SRAR-based output metric of the weakly coupled resonant system can maintain relatively excellent linearity under various damping environments,and possesses superior non-ideal error rejection.The output characteristics of the weakly coupled resonant system in terms of input harmonic stiffness perturbation are studied,and it is verified that the weakly coupled resonator based on the SRAR-based output metric can be applied to the linear sensing of harmonic signals.The specific sensitive structure scheme of the weakly coupled resonant MEMS bionic hair sensor is designed.The key parameters of weakly coupled double-ended fixed tuning forks(DETFs),micro-leverage,swing suppression and other structures are analyzed and optimized to improve the mechanical sensing performance of the sensor.The finite element method(FEM)simulation model of the hair sensor is established.Based on ANSYS software,the following FEM simulations are implemented: modal simulation,thermal stress simulation,transient response simulation,harmonic response simulation,fluid-structure coupling simulation,acceleration response simulation,cross-axis coupling error simulation,etc.The simulation results verify that the performance of the designed weakly coupled resonant MEMS bionic hair sensor satisfies the application requirements.The measurement and control scheme of the weakly coupled resonant MEMS bionic hair sensor is studied.The closed-loop drive circuit based on the phase-locked loop(PLL)and the signal detection circuit based on the C/V converter are designed,and the implementation scheme of the digital measurement and control system based on field-programmable gate array(FPGA)is determined.The signal demodulation scheme based on quadrature coherent algorithm is studied for the extraction of the harmonic stiffness perturbation signal,and the effectiveness of the demodulation scheme is verified by the SIMULINK system simulation.A direct measurement scheme of amplitude ratios is proposed,and the effectiveness of the scheme is theoretically verified based on the averaging method and SIMULINK system simulation.The fabrication and experiments of the weakly coupled resonant MEMS bionic hair sensor are implemented.Based on the deep dry silicon on glass(DDSOG)and micro-assembly process,the prototype of the weakly coupled resonant MEMS bionic hair sensor is fabricated.A test platform is built in the laboratory and the performance tests of the hair sensor prototype are implemented.The experiment results demonstrate that the tested weakly coupled resonant MEMS bionic hair sensor prototype can effectively detect the external steady-state airflow velocity and oscillating airflow velocity,and the SRAR-based output metric can significantly improve the mechanical sensitivity and output linearity of the weakly coupled resonators.In addition,based on different amplitude ratio measurement schemes,the output comparison experimemt of the sensor prototype is implemented.The effectiveness of the direct measurement scheme of amplitude ratio based on amplitude closed-loop control is finally verified.