Error Compensation Algorithm and Calibration Experiment Based on Ellipse Model for Magnetic Compass in Land Navigation System
Ping Ye,Gang Du,Chuanrun Zhai,Haowen Wu,Xingqun Zhan
2009-01-01
Abstract:The present land navigation system is composed mainly of IMU (Inertial Measurements Units) and some other auxiliary sensors. Although IMU can give all-round behavior representations using angle velocity and acceleration, it is inevitable that the measurement error, angle in particular, increases with time rapidly. Based on an overall consideration, magnetic compass, especially for its low cost, small size and low power loss, is generally selected as an important auxiliary sensor and popular with integrated system to eliminate the angle error.When magnetic compass is configured in land integrated navigation system, it can give heading information as an auxiliary attitude reference sensor. The precision of magnetic compass serves an important role in the performance of integrated system. Since magnetic compass measurements are vulnerable to magnetism disturbances from surrounding iron magnetic materials, the calculated heading is distorted as well. In actual application, the heading distortion may reach up to ten degrees. Correspondingly, the performance of integrated system can surely meet with degradation. Therefore, magnetic compass measurements error compensation is necessary and vital before giving the reference attitude.In this paper, three aspects are highlighted. Firstly, magnetic measurements error tests under different magnetic disturbances are developed, and then heading distortion calculated from the magnetic measurements is analyzed. Secondly, according to the characteristics of the measurements, an error model based on ellipse hypothesis is proposed. Considering low calculations and high stability, direct least square fitting is applied in ellipse curve fitting. Finally, magnetic compass is equipped on vehicle to conduct the land experiment, and GPS attitude determining system with high precision is selected as an attitude reference system. The experimental results have demonstrated that the magnetic compensation algorithm is effective and convenient in land navigation system.