Study On Microwave Tomography With Metallic Boundary Conditon

Microwave tomography is one of the important issues concerned by people in microwave theory, it can be used in health care, military, industry and so on. Microwave tomography is a kind of microwave inverse problem and it is a typical ill-posed problem. Sort of means have been introduced to deal with microwave tomography problem, though there still have two main diffculties: Firstly, low signal to noise ratio caused by interference in the area and boundary condition modeling error may lead to imaging failure. Secondly, the shortage in information obtained by system about object of interest may lead to imaging of low quality or imaging failure. In order to solve the problems mentioned above, this thesis aims on two aspects: Firstly, microwave tomography with metallic boundary condition which can provide a certain boundary condition to be modeled. Secondly, using inhomogeneous background to enhance microwave tomography and obtain more information about object of interest.The current researches on microwave tomography were focused on systems with homogeneous backgrounds in free space, while the work of this thesis was focused on the systems in metallic chamber with inhomogeneous background, the fields of this technique was extended. Further more, microwave tomography systems bedded in a metallic chamber have some advantages over the ones in free space, they can provide a more accurate boundary condition and prevent interferences from surroundings. Aspects about embedded systems were studied, include the systems with different numbers and locations of transmit/receive antennas, the selection of working frequency and the effects of symmetrical systems.In order to enhance the capabilities of microwave tomography system, an inhomogeneous background was employed. We placed a known object, which the complex permittivity and the shape were known, in measurement domain, the known object was considered as prior information and the information was used in microwave tomography algorithm to improve the final results. A metric which based on the singular value spectrum of integral operator was introduced with the purpose of evaluating the inhomogeneous background. The metric can evaluate the capabilities of microwave tomography system by one number, so it can be combined with general optimization algorithms to complete an optimization design. Then an layered optimization inhomogeneous background was designed using simulated annealing method.Synthetic studies were employed and the results show that the inhomogeneous background can enhance the ability of microwave tomography system. There is a strong correlation between the metric and the imaging errors, which were used to evaluate the system and evaluate the imaging results, respectively. The optimized inhomogeneous background which designed base on the metric can significantly improve the microwave tomography imaging accuracy and the convergence rate, the imaging results obtained with the optimized inhomogeneous background were better than the ones obtained with free space.