Fast Solution Of RCS Of Conducting Targets Based On Improved CBFM

Radar cross section(RCS),as an important parameter for acquiring information of the target,has been a key part in research of computational electromagnetics,and it is widely used for target detection,microwave radio frequency and geological survey.In recent years,various numerical algorithms have been proposed around how to accurately and quickly acquire the RCS of complex target,among which the method of moments(MOM)is a basic theory of this research.Due to the MOM will produce a full rank matrix by dispersing integral equation,which lead to long time calculation and high memory consumption.The characteristic basis function method(CBFM)is a macro basis function method which has been developed based on MOM.This thesis improves the traditional CBFM and proposes the following methods to quickly analyze radar cross section of conducting targets.The first method is fast analysis of bistatic radar cross section of conducting targets using multilevel characteristic basis function method(MLCBFM)and recompressed adaptive cross approximation algorithm.In the implementation process of CBFM,in order to solve the matrix equation and improve the speed of calculation effectively,the target is usually divided into larger blocks,so the number of unknowns in each block is correspondingly increased,the calculation of characteristic basis function(CBFs)will become time consuming.In order to solve this problem,the MLCBFM is used to divide the geometry of target into multiple levels,and the process of generating CBFs is decomposed level by level to further compress the unknowns.Besides,the RACA algorithm is used to accelerate the impedance matrices of far field at each level,which can accelerate the calculation of the reduced matrix and generation of the secondary characteristic basis functions(SCBFs),and the computational memory can be further reduced.The second method is fast calculation of monostatic radar cross section of conducting targets using hierarchical characteristic basis function method(HCBFM)and singular value decomposition.For the traditional singular value decomposition-characteristic basis function method(SVD-CBFM),it need the large number of incident excitations and the calculation of the CBFs are time consuming,so the hierarchical CBFM is proposed to solve this problem.In HCBFM,the target is firstly divided with hierarchical partitioning approach,in order to reduce the number of incident plane waves,the CBFs defined in large blocks are expressed as a linear combination of the previously generated low-level primary characteristic basis functions(PCBFs)and SCBFs defined in the relatively small blocks,which can avoid directly solving the CBFs in the large blocks.Finally,the high-level CBFs in large blocks are orthogonalized by using singular value decomposition(SVD)at multiple excitations,a set of linearly independent CBFs can be obtained.Besides,the ACA algorithm is also employed to compress the impedance matrix of far field,which can further speed up the process of calculation.Several numerical examples are analyzed in this thesis to verify high efficiency of the two methods and a new solution is provided for solving electromagnetic scattering problem of complex electric large targets in the future.