Characterization and Calibration of Measurement Error Associated With Attitude Drift of a Coil Vector Magnetometer

Abstract

In long-term observations, the attitude drift of a coil vector magnetometer caused by mechanical deformation will directly affect the accuracy of magnetic direction measurements. To solve this problem, we propose a characterization model and calibration method for the measurement error associated with attitude drift. The proposed model uses an attitude matrix to define a mapping from the unmeasurable characteristic angle to the measurable magnetometer attitude. We propose a new <i>dif</i> coordinate system to establish a characterization model linking the characteristic angle of the magnetometer attitude to the measurement error of the magnetic direction, thereby permitting quantitative analysis of the attitude drift and real-time suppression of its negative influence on the measurement accuracy without complex manual operations. The proposed model shows that in the presence of small attitude changes, the measurement error of the magnetic inclination is mainly determined by the south-north non-levelness of magnetometer, and the error of magnetic declination is mainly determined by the magnetometer orientation and east-west non-levelness. Our simulations show that when the east-west non-levelness, south-north non-levelness, and orientation angle of the magnetometer are 232.6&#x2033;, 206.2&#x2033;, and 615.4&#x2033;, respectively, the measurement errors (baseline shift) of the magnetic inclination and declination are 206.9&#x2033; and 364.2&#x2033;, respectively. To verify the results, we designed an experimental platform and conducted a comparative experiment. The results show that in the presence of the aforementioned attitude parameters, the measurement errors of the magnetic inclination and declination can be reduced by 97.8% and 88.6%, respectively, through error calculation and baseline correction.