An Improved VMD Method for MGTS Calibration and Target Tracking
Qingzhu Li, Jing Li, Zhiyong Shi, Zhining Li, Xue-zhong Wen · IEEE Sensors Journal · 2023
The magnetic gradient tensor system (MGTS) is disturbed by the measurement noise and electromagnetic field, which affects its accuracy of calibration and target tracking seriously. Hence, we proposed an improved variational modal decomposition (IVMD) noise-reduction method for the measurement and calibration of MGTS. First, continuously sampling around the center of MGTS to obtain a sufficient amount of three-axis magnetic component signals, and the total magnetic intensity (TMI) and cosine signals are vector-synthesized to avoid distortion. Then, the endpoint extension of the signal is performed using the spectral cyclic correlation coefficient, and variational modal decomposition (VMD) on the synthesized signal and kernel independent component analysis (KICA) are performed on its principal components to eliminate interference noise and modal aliasing. Finally, the system is calibrated with the integrated compensation and rotation alignment (ICRA) method to eliminate sensor systemic error, array misalignment error, and hard or soft magnetic interference. Experiments show that, compared with ICRA direct calibration, the proposed method reduces the root-mean-square error (RMSE) of the MGT component from 102.2 to 5.6 nT/m in the indoor electromagnetic field environment. In this scene, the tracking accuracy RMSE of the magnet target is reduced from 0.16 to 0.07 m.