一种基于输出误差观测的冗余MEMS-IMU标定技术研究

针对传统六位置标定存在标定步骤复杂、标定时间长等问题，同时随着MEMS器件冗余数目增加和冗余配置结构复杂化，利用分立标定技术实现器件误差参数辨识的难度进一步增大，且不同配置结构所采用的标定方法存在通用性较差的问题。因此提出了一种基于Kalman滤波的冗余MEMS-IMU分立标定方案。该方案首先采用小角度建模法实现安装误差的精确建模；然后针对直接以转台三轴角速率为观测值，导致部分状态量不可观的问题，提出了以器件的输出误差值作为观测量、以器件误差参数作为状态量设计Kalman滤波器；最后设计了高精度三轴转台转位编排方式，并利用四陀螺冗余结构进行标定仿真试验。仿真结果表明：该标定方法与六位置标定方案相比，标定精度平均提高了11.37%，可实现MEMS器件误差参数的快速辨识，对实际工程实践具有一定参考价值。

Research on Redundant MEMS-IMU Calibration Technology Based on Output Error Observation

As the traditional 6-position calibration has complicated calibration steps and long calibration time, and with the increase of the number of redundant MEMS devices and the complexity of redundant configuration structure, it is more difficult to use discrete calibration technology to realize the error coefficients identification of devices. Besides, the calibration methods adopted by different configuration structures have poor generality. A redundant MEMS-IMU discrete calibration scheme based on Kalman filter is proposed in this paper. The small angle modeling method is firstly used to realize the accurate modeling of the installation errors. Then, because the triaxial angular rates of the turntable are directly used to observations, which makes part state variables unobservable, the output errors of the device are taken as the observations, and the device error coefficients are used as the state variables to design the Kalman filter. Finally, the high-precision three-axis rotary table indexing method is designed, and a four-gyro redundant structure is used for the calibration simulation. The simulation results show that compared with the six-position calibration scheme, the calibration accuracy is improved by 11.37% on average. The fast identification of MEMS device error coefficients is realized, and the scheme has certain reference value for practical engineering practice.