Characteristic Basis Function Method Based on Singular Value Decomposition and Its Applications in Electromagnetic Scattering Problem
|School||Nanjing University of Aeronautics and Astronautics|
|Course||Electromagnetic Field and Microwave Technology|
|Keywords||Surface Integral Equation Singular Value Decomposition Physical Optics Characteristic Basic Function Method Ideal Conductor Electromagnetic Scattering|
The high efficient and precision electromagnetic numerical method which is used to analysis electromagnetic scattering characteristics of 3D complex structure large electrical conductor, medium body as well as conductors and media mix has become one of the hot topic which scholars at home and abroad to study.In this background, this paper studies characteristic basic function method (CBFM) which based on Moment of Method (MOM), and proposes a new CBFM based on singular value decomposition (SVD)——SVD-POCBFM. This method is based on the traditional CBFM, and its primary characteristic basic function (PCBF) is constructed by a series of planar spectrum of the incident angle space through Physical Optics (PO) method. Then using SVD for these PCBFs, the independent orthogonal basis function group which is irrespective of angle, polarization and/or frequency can be obtained. Then the linear combination of this orthogonal function group is used to state the surface current distribution of metal objects, so as to the quick computation of monotonic characteristic and broadband characteristic of RCS of metal objects can be achieved.The advantages of this paper are:①The CBF by this method contains the information of the incident Angle, polarization and/or frequency, so when the monotonic characteristic and broadband characteristic of RCS is calculating, CBFM is no longer needed to calculate one by one (angle and frequency). Thus computation efficiency will be increased obviously.②When PCBF is calculating, PO method will be adopted, for this reason, this method is not only simple and quick calculation in program, but also not occupy memory.③Because the number of CBFM which contained the information of angle, polarization and/or frequency, ignoring the high CBF is almost no influence for the precision of RCS solution.④Multi-layer SVD has less memory than direct SVD, and computation efficiency is also improved.