Fuzzy Adaptive Extended Kalman Filter for miniature Attitude and Heading Reference System

In the paper a newly developed fuzzy adaptive Kalman filter (FAKF) algorithm is presented which is applied in miniature attitude and heading reference system (AHRS) based on MIMU/magnetometers. The method is to deal with time variable statistic of measurement noise in different working conditions. By monitoring the innovation of sensors data in realtime, the Kalman filter tunes the measurement noise covariance matrix and process noise covariance matrix on-line according to fuzzy logic inference system to get the optimal state estimation. The test results indicate that the algorithm of FAKF has better accuracy than the regular Kalman Filter.