High accuracy extend Kalman filter for posture measurement based on attitude and heading reference system

Due to the drift of gyroscope, there is lots of accumulated error in the posture measurement. The gyroscope can't complete a long time and high accurate attitude measurement independently. Therefore, the common method is the multi-sensor information fusion technology. The measurement system based on gyroscope, accelerometer and magnetometer is presented by this paper. The original data is compensated and filtered by extend Kalman filter. And the sensornoise is reduced. The Kalman state equation is established by the gyroscope data calculated by quaternion. The Kalman measurement equation is established by the output of accelerometer and magnetometer. The experiments' results show that the error of the attitude angle is effectively suppressed by this extend Kalman filter.