Magnetic Interference Data Evaluation and Compensation of Three-Axis Magnetometers

Three-axis magnetometers (TAMs) provide rich measurement information and are widely used in magnetic target detection, geological measurement, geomagnetic navigation, and other fields. The interference from the TAM itself and carriers significantly affects TAM measurement and detection performance and, thus, needs to be compensated before use. Insufficient data often lead to compensation failure. How to evaluate the relationship between effective data volume and compensation performance has not been solved yet. To address this issue, we propose a compensation data evaluation method. By constructing a data matrix related to compensation data, the distribution of the matrix rank is analyzed to determine the number of valid singular values that can be used for compensation. The results indicate that when the TAM rotates around a single axis, two axes, and three axes during compensation; the number of valid singular values is five, eight, and nine, respectively. When the TAM rotates freely during compensation, the number of valid singular values is related to the inflection point of the singular value distribution. When the number of valid singular values is applied to interference compensation and target detection, good results are achieved. Before compensation, magnetic noise (peak-to-peak) caused by the TAM rotating with the seawater is 235.2 nT, and the target signal is submerged in the noise and cannot be detected. When the rank of 9 is used for compensation, the compensated value deviates from the true value and cannot be used for detection. When the rank of 8 is used, the noise after compensation is 8.6 nT (peak-to-peak), and the magnetic anomaly target can be easily detected, which is 21.6 nT (peak-to-peak).