Complex Target Electromagnetic Modeling And Pre Conditions To Accelerate The Technology Research

Nowadays, more and more attention is paid to the analysis of radar cross section(RCS) in modern military fields. The most of the simple targets have been analysed perfectly in the past decades. It becomes more and more necessary to improve the technique of modeling with the development of the modern military. In this thesis, a mature modeling technique based on Spline surface is introduced to rebuild some models, such as F22. All kinds of the necessary electromagnetic meshes can be easily obtained from the rebuilded models.This thesis is based on the method of moment (MoM) which has been widely used in the analysis of the electromagnetic scattering problems. However, the MoM costs a lot of memory and time for electrically large object. It is too difficult to solve the large dense linear systems fast and accurately, which formed by the integral equations.The parallel the multilevel fast multipole method (MLFMM), which is based on the computing platform of MPI, is introduced to accelerate the matrix-vector product operation.Although the MLFMM and the parallel technology can accelerate the matrix-vector product operation, the number of iterations does not reduce at all in the iterative solution. Therefore, the preconditioning technology becomes more and more popular for the complex targets. The parallel sparse inverse preconditioning(SAI) based on MLFMM is introduced to effectively reduce the overall simulation time. Moreover, the parallel SAI used in EFIE has been expanded to CFIF in this thesis. The parallel MUMPS preconditioner is used for large and complex targets. In addition, a kind of two-step preconditioning technique is presented to further reduce the iteration numbers. Numerical experiments are presented to show the efficiency of the spectral two-step preconditioner.