MEMS陀螺误差辨识与补偿

Error identification and compensation of MEMS gyroscope

由于制造工艺等原因,MEMS陀螺的随机漂移非常大,严重影响了系统的性能。通过自制的基于MEMS的捷联惯导系统的相关实验,对MEMS陀螺的确定性误差和随机误差分别进行了辨识和补偿。完成确定性误差补偿,对MEMS陀螺随机误差进行了时间序列分析,并建立了AR模型,根据所选模型参数建立了随机误差的系统方程,采用经典卡尔曼滤波进行随机误差补偿。实验结果说明:无论是静态下还是动态下,补偿后信号的方差都大大下降,说明了滤波效果较为明显,具有一定的工程应用价值。

MEMS gyro＇ s random drift is very large, because of the manufacturing process and other reasons, which seriously affect on system performance. Through experiments of MEMS strapdown inertial navigation system, deterministic and stochastic error was identified and compensated. After deterministic error was compensated, the stochastic error was analyzed based on time series and AR model was set up. System equation of stochastic error was established based on the selected model and the error was compensated by Kalman Filter. Test results show that variance of MEMS gyroscope stochastic error reduced greatly after filter, which illustrates the filtering effect is obvious,and has a certain value of engineering application.