Research On Scattering Center Extraction And RCS Extrapolation Of Metal/thin Coating Targets Based On GTD Mode

Under the background of modern military electronic countermeasures,the research of radar target characteristics can provide theoretical guidance for target stealth design and recognition.As one of the important contents in the research of radar target characteristics,the establishment of the parametric model of scattering center is more convenient for the research of scattering characteristics.In the actual observation process,part of radar echo data is often lost due to faults.Compressed sensing can be used as an efficient tool to solve the problem of sparse signal recovery.The scattering center extraction of metal and thin coated target GTD model and the application of RCS extrapolation from its parameterized model based on compressed sensing are studied by this thesis.The estimation performance of the compressed sensing algorithm is implemented well only when the scattering center is assumed to be located on the grid.However,the scattering centers on the radar target are continuously distributed,so the scattering centers on the grid can not all be ensured.And the estimation accuracy can be affected when the scattering points are not be located on the grid.For the compressed sensing off-grid problem,two scattering center extraction algorithms based on disturbed dictionary matrix for metal/thin coated target GTD model proposed by this thesis.One is orthogonal matching tracking algorithm based on disturbance dictionary.The most correlated atom can be selected by this method.Then the nonlinear programming function is applied to solve the optimal grid offset.The optimal parameter can be obtained by iterating the above process.The second one is the sparse Bayesian algorithm based on disturbance dictionary.The GTD model is expanded by Taylor series at grid points.It is noted that the distribution of the grid offset is uniform.The sparse Bayesian principle is used to get parameter values with maximum posterior probability of signals.Finally,the RCS extrapolation of metal and thin coated target with different size based on one-dimensional scattering center parameterization model is proposed.It can be used for the measurement of electromagnetic scattering parameters of large radar targets.In order to enhance the estimation precision and decrease the solving time,an algorithm of scattering center extraction based on one-dimensional continuous compressive sensing for metal/thin coated target GTD model is proposed in this thesis.This method eliminates the need for discrete grids.In this way,the atomic norm minimization can be converted to semi-definite programming problem.The two-dimensional GTD-based scattering center extraction is proposed for metal/thin coated targets to describe the azimuth angles.The correlative alternating direction multiplier method is also given to accelerate the solving process of one-dimensional and two-dimensional semi-definite programming problems.Finally,the RCS extrapolation of metal/thin coated target with different size based on two-dimensional scattering center parameterization model is proposed.Finally the efficient analysis and extrapolation application based on scattering center extraction of radar target are realized.