Susceptibility inversion of near-field magnetic sources and its application

Abstract

Susceptibility inversion of near-field magnetic sources is of great application significance in the field of magnetic target detections, such as mine clearances and buried pipeline detections. This paper proposes a method of three-dimensional inversion imaging of magnetic sources in near-field with unknown shape, position and magnetization state. The method transforms the total field norm anomaly ΔT on the observation surface into the total field vector anomaly ||Ta|| and then constructs an objective function aiming at minimizing the deviations between the measured and predicted magnetic anomalies. The magnetic dipole equivalence principle is utilized to calculate the kernel matrix. The Gauss-Newton and conjugate gradient algorithms are utilized together to solve the inversion optimization process. The range of the magnetic susceptibility is constrained by using nonlinear transformations. Simulations and measurement experiments demonstrate the proposed method is capable to realize 3D inversions of curved pipes, intersecting pipes, separated magnetic sources, and “L” shaped steel pipes under near-field conditions. Through the actual magnetic field measurement experiments and the implementation of the proposed optimization inversion algorithm, accurate inversion of the magnetic susceptibility distribution and magnetic imaging of the “L” shaped steel tube was realized, testifying that the proposed method is feasible to be applied to the detection of buried pipelines.