Magnetic target nonlinear positioning method based on planar regular octagon tensor system

Abstract The magnetic target location technology based on the principle of magnetic gradient tensor has broad engineering application prospects. However, commonly used target positioning methods are easily interfered with by the geomagnetic environment. We proposed a magnetic target nonlinear positioning method to achieve precise localization of magnetic objects. An ellipsoid fitting method based on the total least square algorithm is proposed to correct the measurement system’s magnetic interference and array error. The results show that the algorithm can reduce the root mean square error of the total magnetic field strength by 93.05% to 0.1007 uT. The tensor components are all limited to 100 nT m −1 . Then, a target positioning function is constructed with the magnetic moment information introduced as a constraint term, and the location of the magnetic target is optimally calculated using the Levenberg–Marquardt algorithm. The field experimental results indicate when the magnetic targets are located at (5,0,0) and (10,−5,0), the positioning errors are 5.11% and 6.62%. This technology can also achieve high-precision path tracking of magnetic targets.