Comparison and Study on Complementary Filter and Kalman Filter in AHRS Attitude Calculation

DING Jun,ZHAO Zhong-hua
DOI: https://doi.org/10.3969/j.issn.1006-9348.2013.09.014
2013-01-01
Abstract:The accelerometer measurements contain the gravitational acceleration and linear acceleration,and the magnetic sensor can be easily disturbed by ferromagnetic substance. So it causes roll and pitch angle errors when using the accelerometer measurements to calculate,and it then directly introduces an error to the heading calculation which uses the magnetic sensor measurements. In order to reduce the error of attitude angles,which are calculated by the attitude calculation method based on the accelerometer and magnetic sensor,this paper designed a complementary filter and a Kalman filter separately to filter the outputs of accelerometer and magnetic sensor using the outputs of gyroscope,used MATLAB to do simulation with the data collected from Micro Inertial Measurement Unit( MIMU) of VN-100,and finally compared the results of the two filters. Simulation results show that both the complementary filter and the Kalman filter can improve the attitude angles accuracy,and the complementary filter is more effective than Kalman filter.
What problem does this paper attempt to address?