Local Magnetic Field Inversion Based on the Modified L-Curve Model Function Method

In recent years, the strategic position of the ocean has become increasingly important. Geomagnetic navigation has become the focus of underwater navigation research due to its advantages of all-day and all-terrain. It is urgent to establish a high-precision local marine geomagnetic field structure. At present, the research methods of local geomagnetic fields mainly focus on data-driven, which leads to the lack of universality, and the limited detection information leads to the inability to achieve high-precision modeling. Therefore, this paper proposes a local magnetic field inversion and reconstruction model-driven method based on the Tikhonov regularization method, building upon the magnetic dipole array model. Then, to improve the modeling accuracy and reduce the computational complexity, the regularization parameters of the Tikhonov regularization method are optimized. We propose a modified L-curve model function method, which can quickly locate the optimal regularization parameter in the case of unknown error levels. In the experimental analysis, the magnetic field measurement data measured after constructing the field model is used as an example. The magnetic moment information obtained by inversion and the magnetic dipole array model is used to reconstruct the local magnetic field model. The reconstruction results provide the overall magnetic field map and the magnetic field intensity data at any given point. Applying this method to marine magnetic field reconstruction, enhancing the accuracy of submarine navigation and improving anti-submarine warfare capabilities, strengthening underwater defense technology capabilities.