Research of on-line modeling and real-time filtering for MEMS gyroscope random noise

With the purpose of improving the performance of MEMS gyroscope, reducing the MEMS gyroscope random noise, a real-time fuzzy adaptive Kalman filter based on time-sequence model is presented in this paper. Considering the output of MEMS gyroscope is weak stationary and nonlinear, the random noise is on-line modeled through an advanced recursive least squared algorithm to amend the time-sequence model in real time, and filtered by a fuzzy adaptive Kalman filter, realizing the real-time adjustment for measurement noise. Compared with the traditional Kalman filter based on off-line built time-series model, the novel method is more competent to reduce the random noise in practical use, the result shows that the new adaptive Kalman filter achieves a higher accuracy than the classical Kalman filter's with strong robustness.