Research on the random error modeling and compensation method for MEMS gyro

Abstract

The random drift of MEMS gyro is large because of external environmental factors and manufacturing technique. The random drift modeling and compensation of MEMS gyro CRS03 is studied. The stability and normality of data about 2minutes, 5minutes and 10minutes long, outputted from the MEMS gyro are analyzed. The research result shows that the outputting data of MEMS gyro is stable of short-time. The time series analysis technique is used to set the ARMA model of MEMS gyro random drift. And the Kalman filter is used to estimate the random drift of MEMS gyro. The research results show that the random drift of MEMS gyro is compensated effectively and the precision of MEMS gyro is improved.