摘要

Magnetic induction tomography (MIT) is an imaging technique based on the measurement of the magnetic field perturbation due to eddy currents induced in conducting objects exposed to an external magnetic excitation field. In MIT, current-carrying coils are used to induce eddy currents in the object and the induced voltages are sensed with the receiving coils. When the driving frequency is significantly high relative to the frequency range in which MIT normally operates, metallic targets with high conductivity between the coils can be treated as perfect electric conductors (PEC) with negligible errors. In this scenario, the penetration depth of the magnetic field into the target is extremely small and the traditional versions of the finite element method (FEM) are not efficient for the calculation of the sensitivity and the forward problem due to the requirement for large number of elements to reach an acceptable computational precision. Other versions of FEMs (such as Hp-FEM), which have higher discretization efficiency and more advanced elements to satisfy the requirement, are exceptions. Nevertheless, the discretization regions for all FEMs have to extend beyond the region that contains the conducting object and volumetric elements are generally required for 3D problems. In contrast, the boundary element method (BEM) based on integral formulations becomes an effective way to analyze this kind of scattering problem since meshes are only required on the surface of the object. By point collocation, the boundary integral equations can be transformed into linear equations. Numerical methods are used to solve the linear equations and the solution of the original integral equations can be obtained. In this paper, we compute four typical sensitivity maps between the coil pairs in high-frequency MIT system due to a PEC perturbation. The magnetic scalar potential is used to improve the efficiency. Five PEC objects of different shapes are used in the simulation. The results have been compared with the experimental results and that obtained from the H center dot H formulations. We can know that the sensitivity maps derived by BEM are in good agreement with that from the experiment and theoretical solution. Overall, BEM is an effective way to calculate the sensitivity distributions of a high-frequency MIT system.

全文