微机械陀螺随机误差补偿研究

Research on Compensating of Random Error of MEMS Gyroscope

为了提高微机械陀螺使用精度,研究了随机漂移误差补偿方法。对静态陀螺输出数据进行预处理,提取出随机漂移数据,采用时序分析方法对其建立AR模型。基于模型设计Kalman滤波器,进行静态和动态仿真。结果表明:静态条件下滤波效果显著;动态条件下,滤波效果变差。针对此问题设计了标量因子时变的自适应Kalman滤波器,试验表明,此算法能够有效降低微机械陀螺的随机漂移。

In order to improve the performance of MEMS gyroscope, the compensating method of random drift error was analyzed and modeled. Based on the time series analysis method, the data of random drift was taken out from the processed original measurement of the MEMS gyroscope, and was modeled lay the AR（1） model on time series analysis method. With the AR model, a Kalman filter was designed. By the tests, it showed that the error was remarkably decreased after fil-tering in the static test, while the effect was not good in the dynamic test. Then a time-vary factor adaptive Kalman filter was designed, and it could effectively decrease the random drift of MEMS gyroscope.