The Study Of Key Technology For Fast Calculation Of Wide-band RCS For Electrically Large Conductor Target

In computational electromagnetics, the study of electromagnetic scattering property for object at a frequency band continuously plays an important role. Because the research has significant real meaning. For example, target identification of ultra-wideband radar, stealth and anti-stealth technology, which all need the bandwidth electromagnetic scattering property of the target. For the fast analyzing of the property, in this paper, based on the traditional method of moments (MoM), works of several aspects are proposed as following.Firstly, the basic principle and related concepts of the MoM are introduced. The detail solving process of the radar cross section of the three-dimensional perfect electric conductor target with the MoM is presented. The asymptotic waveform evaluation (AWE) technique is also introduced, which has been applied widely to the computation of wide-band RCS for target. The wide-band RCS of several simple targets are computed with the AWE technique.Then, Krylov subspace method is introduced, and several popular iterative methods, such as BiCG, BiCGSTAB, GMRES, are used to solve the large matrix equation originated from the discretization of electric field integral equation. Their performances will be compared through the numerical results from different examples. In order to improve the convergence rate of iterative method, the preconditioning technique is elaborated. The inner-outer iterative near field preconditioning technique combined with the AWE technique improve greatly the computational efficiency of wide-band RCS for target.The least square fitting (LSF) method, which is originally used to the curve fitting, is first introduced into the study of bandwidth electromagnetic scattering property for object. Compared with the AWE technique, the LSF method needs less memory in computational process. Thus, the method can improve computational efficiency without sacrificing any accuracy and may be applied to the computation of wide-band RCS for electrically large target. Finally, the parallel computing is studied in this thesis. Under the Windows operating system, a cluster computing system is built using personal computer. The message passing interface (MPI), which has been widely utilized in parallel computing, will be used as parallel environment. The AWE technique will be paralleled first in this thesis. The numerical examples show that the AWE technique combined with the parallel computing can reduce significantly the computing time of wide-band RCS for electric large target.