Research On The Inverse Problem Of Electromagnetic Tomography And Sensing Technology

For the advantages of high speed acquisition, non-contact, radiationless, simplestructure and low cost, Electromagnetic Tomography and Sensing Technology(EMTAST) has broad prospects in industrial process parameter detection andbiomedical field.This thesis aims at a systematical study on the inverse problem of EMTAST;achievements and progress have also been made in the priliminary applications of thistechnology in practice. The main contents of the thesis are summarized as follows:The Finite Element Method (FEM) was applied in the computation of scalarpotential and thus the sensitivity matrix was obtained from simulation. The FEMoffered certain reference frame for the sensitivity measurement. The sensitivity matrixwas obtained by perturbation method to reconstruct the images. A new criterion forevaluating the uncertainty of images reconstructed by different simple algorithms wasproposed. The new criterion was the singal to noise ratio of pixels of reconstructedimage.The traditional non-iterative algorithms and iterative algorithms were studiedrespectively. A modified Tikhonov regularization algorithm was proposed, whichweaken the excessive smooth funcition of the results compared to the standardTikhonov method. To deal with the slow rate of convergence of iterative algorithms, apre-iterative method and preconditioning approach for SIRT (simultaneous iterativereconstruction algorithm) were proposed and proved by experiments that theimproved methods had well real-time performance and satisfactory image quality.The application of Electromagnetic Sensing Technology in the thichknessdetection of nanoscale metal film was proposed for the first time. By the analyticalsolution and FEM simulation of a simplified EMT model, the issues on how to choosethe excitation frequency and measurement basis were explained in detail. The EMTsystem were modified and used for the thickness measurement, where the results werewell comparative with the analytical solution and FEM simulation. The consistence ofthe results showed that Electromagnetic Sensing Technology could be applied in thethickness measurement of nanoscale metal film.Because of the advantages of light, stout, antifatigue and anticaustic, CFRParoused wide concern in aerospace and motor industry. This thesis tried to estimate the conducting direction of CFRP in a creative way. Beginning with a simplifiedanalytical model, a method to evaluate the equivalent conductivity of CFRP wasexplained. As a result of the anisotropy of CFRP, the FEM simulation could be used tomeasure the conducting direction and overlay sequence of the single-layer ormultilayer CFRP. Experiments were conducted with modified EMT system and theresult showed well agreement between the experimental result and FEM simulation,which proved that Electromagnetic Sensing Technology is an effective method fornondestructive testing of CFRP.Finally, the research work was briefly summarized; suggestions and advices forthe further development of EMTAST are presented as well.