Online estimation method of Allan variance coefficients for MEMS IMU

As a noise analysis of MEMS IMU, the traditional Allan variance methods have large computational burden because of requiring to store a large amount of data. Moreover, the procedure of drawing slope lines for estimation is also painful. In order to overcome these drawbacks, a online method is proposed to estimate the Allan variance parameters, which directly model sensors random errors including quantization noise, angular random walk, bias instability, rate random walk and rate ramp into a nonlinear state space model and then implemented by sage-husa adaptive Kalman filter algorithm. The comparison of results of real ADIS16405 IMU static gyro noise analyzed by Allan variance method and the proposed approach shows that the results from the proposed method are well within the error limits of Allan variance method. Moreover, the technique proposed here estimates the Allan variance coefficients in real time, effectively avoids storage of history data and manual analysis for an Allan variance graph