硅MEMS陀螺驱动/检测模态频率调谐自适应控制方法和实验研究

Resonant frequency difference between driving and sensing modes are one of the most critical parameters that influences the sensitivity and linearity of silicon MEMS gyro. Other than improving the modes symmetry by manufacturing technics there is a more creative and economical way to suppress the resonant frequency difference between driving and sensing modes. Precise modeling, resonant frequency tuning control by micro electrodes, adaptive control to suppress the frequency drift caused by temperature will be studied in this project. 1) Ring-structure silicon MEMS gyro is chose for its abundant electrodes and good symmetry as well as controllability. Micro mechanics and system identification by genetic algorithm are combined to get more precise dynamic models of ring-structure silicon MEMS gyro. 2) Resonant frequency difference between driving and sensing modes is the aim of optimal control. The control voltages on those tuning electrodes are determined by a modified genetic algorithm and simulated annealing mechanism. 3) In order to suppress the frequency drift caused by temperature adaptive feed-forward and feedback control are applied with Lyapunov function to enhance both the robustness and stable precision of our tuning methods. True data will be used in modeling and frequency tuning-control; controlled gyro will be tested in the practical working environment; all our work will be done on the basis of the previously established research platform of driving and sensing technology. Our work will contribute to improve the performance and rate of completed products of silicon MEMS gyro.

驱动/检测模态的频率调谐特性对硅MEMS陀螺灵敏度和线性度有显著影响。除不断改进加工工艺提升模态对称性，运用自适应最优控制理论、通过建模-调谐控制-温度漂移抑制等三步骤使硅MEMS陀螺达到并保持调谐状态是更高性价比的技术途径。本项目1)针对电极配置灵活、可观性可控性强的环状硅MEMS陀螺，通过微机械力学分析和遗传优化算法精确建模；2)以模态谐振频率差为优化目标，以调谐电极偏置电压为控制手段，采用改进遗传算法结合模拟退火算法确定调谐控制的最优电压配置；3)引入带Lyapunov函数的自适应前馈反馈控制器抑制温度变化引起的驱动/检测模态频率失谐，提高动态响应速度和稳态跟踪精度。本项目基于实测数据完成建模和调谐控制，并通过同等条件实验验证陀螺驱动/检测模态谐振频率差和稳定性。结合现有驱动/检测技术基础和研究条件，实现实际环境下环状硅MEMS陀螺的调谐控制，将显著提升硅MEMS陀螺性能和成品率。

驱动/检测模态的频率匹配特性对硅MEMS陀螺的性能有显著影响。除了采用工艺手段提升两个模态对称性，将自适应最优控制方法与硅MEMS陀螺的设计加工紧密结合，通过建模、调谐控制、温度漂移抑制等手段使硅MEMS陀螺达到并（在实际使用环境中）保持调谐状态是亟待开发的技术途径。本项目针对电极配置灵活、可观性可控性强的环状硅MEMS陀螺，通过实验测试和遗传算法优化参数、建立了准确动力学模型，为进一步调谐控制提供理论基础；以驱动/检测模态的谐振频率差最小化为控制目标，以调谐控制电极偏置电压为控制手段，将调谐控制转化为模型参数优化问题，采用改进的遗传算法结合模拟退火算法给出了调谐控制电极的最优配置，并通过实验验证调谐后的驱动/检测模态谐振频率差；抑制温度变化引起的频率失谐中引入带Lyapunov函数的自适应前馈反馈控制器，提高了动态响应速度和稳态控制精度，保证了抑制算法的精度和温度跟随性。. 本项目成果包括：研制硅MEMS陀螺测试修调专用设备样机1台；在航天九院十三所和航天九院772所得到应用；环状硅MEMS陀螺修调后全温范围-40℃～60℃误差缩小到28%以下；发表SCI检索论文4篇，EI检索论文3篇；申请发明专利5项，获批软件著作权3项；获得国防技术发明二等奖1项，排名第一。