Identification and elimination of outliers in magnetic heading information measurement

Abstract

The magnetic sensor is easily disturbed by the dynamic magnetic interference and leads to the magnetic heading measurement error when the carrier moves or flies near the ground. This paper puts forward a method which uses the federal Kalman filter for identification and elimination of outliers in magnetic heading information measurement. The influence of outliers on the results of the ordinary Kalman filter output was analyzed. A carrier heading measurement method based on federal filter structure was proposed. The method for identification and elimination of the magnetic sensor outliers was designed by using a two-stage federal Kalman filter. The information fusion algorithm of the master-filter and sub-filters was designed. The calculation method of the information distribution factors for the sub-filters was studied. Experiments proved that the method can effectively identify and eliminate outliers of the magnetic heading information.