The Study On Microwave Imaging Of Two-dimensional Perfectly Conducting Objects

Microwave imaging is a traditional imaging technology. Microwave is the information carrier. Both amplitude and phase of the measured scattering data can be used in microwave imaging. It has been studied extensively in recent years due to considerable demand for practical applications in target identification, subsurface and ground-penetration radar, geophysical remote sensing, medical imaging etc. It is a highly nonlinear and ill-posed problem. In the past, direct methods are widely applied. However, most of these methods have a fatal limitation that they converge to the true profile quickly when a good initial guess is obtained in advance. Otherwise, they may be trapped into a local extreme or even diverge.In this paper, the algorithms of microwave imaging of two-dimensional perfectly conducting objects are mainly discussed. The traditional microwave imaging technology - Newton-Kantorovitch is considered first so as to find the imaging mechanism of perfectly conducting objects, then the genetic algorithm(GA) is induced. GA is a slowly converging probabilistic global optimization method based on genetic recombination and evolution in nature. It operates on a population in the search space simultaneously and performs a global optimization by the three genetic operations, namely, selection, crossover and mutation. GA is less prone to converge to a local optimum even when the initial guess is far away from the exact solution. In recent years, a growing number of researchers in the GA community turn to the study of real-coded genetic algorithm (RGA) for its simplicity and efficiency, and the reason that a chromosome can be directly represented by real number. In frequency domain, the direct problem - electromagnetic scattering problem is solved by the method of moment, and the inverse problem by RGA. This method is called microwave imaging technique based on RGA under frequency domain. After comprehending the principle of the microwave imaging and RGA, we present a novel imaging method for the first time. It is a full-time-domain algorithm based on RGA, FDTD technique is applied to solve direct problem while the RGA is applied to solve inverse problem. Inorder to validate the feasibility of the novel algorithm, ultrawide band pulse wave is used as incident wave. Computational domains are limited by absorbing boundary condition. The transformation technique from near zone to far zone is used when the sampling spot is outside of the computational domains. The performances of the two algorithms (time domain and frequency domain) are examined by computer simulation. Furthermore the time-domain algorithm is also proved by actual experiment.Another contribution in this paper is that a novel absorbing boundary condition based on the one-way wave equations has been proposed. It is more effective both in theoretical analyses and numerical simulations when compared to other widely used boundary conditions such as second order Mur ABC's.