Research On Electromagnetic Scattering Characteristics Of Complex Linear Structure Based On Compressive Sensing

In recent years,with the rapid development of wireless communication narrow-pulse radar technology and electromagnetic compatibility(EMC/EMI),the problems of electromagnetic scattering and radiation of thin-line antennas and linear structures have attracted a great deal of attention owing to their wide applications.At present,the method of moment(MOM)is the main method to analyze the electromagnetic scattering problem,and it is widely used because of its accurate calculation accuracy.However,the traditional MOM has the problem of storage and operation of dense impedance matrix.In this paper,a series of optimization algorithms are proposed for the shortcomings of traditional MOM in solving the electromagnetic scattering problem of linear structures.On the basis of the MOM combined with the compressed sensing(CS),the sparse transformation and the fixed row extraction are used to form a underdeterminted equation model,and the recover algorithms are used to reconstruct the induced current.Applying the proposed method to analysis of three-dimensional wire antenna model,by analyzing the theoretical characteristics of the compressed sensing,we can see that the new method can not only reduce the storage capacity,but also reduce the computational complexity.The main work of the thesis is:First of all,the theoretical framework of the MOM is studied,including the basis function and the weight function,and the division and calculation of the model Meanwhile,the theoretical framework of compressed sensing is introduced eapecially about the key techniques of sparse representation,observation matrix and recovery algorithmSecondly,a fast solution model of CS combined with MOM is established.Elements in the excitation matrix and the corresponding row in impedance matrix are extracted based on sparse transformation,with the help of a reasonable threshold,an underdetermined equation is constructed,and then the induced current can be reconstructed.Thirdly,based on the above fast solution method,different sparse transforms and recovery algorithms are applied to the model and studied.The discrete cosine transform basis(DCT)is compared with the discrete Fourier transform basis(DFT)The matching pursuit algorithm(OMP)and the generalized orthogonal matching pursuit algorithm(GOMP)are compared to find the optimal algorithm,which can reduce the computational complexity and improve the computational efficiency and lay the foundation for the follow-up workFinally,aiming at the electromagnetic scattering problems of wire antennas over a wide angle,a new excitation source is constructed based on CS.The random extraction method is used to form an underdetermined equation under the new excitation source,and it can be solved by recovery algorithms and the current at all incident angles can by reconstructed with high efficiency and accuracy.By comparing with the reconstruction eff-iciency of different recovery algorithms,the optimal algorithm is obtained.Numerical experiments have shown that the algorithm can significantly reduce the computational complexity and with high accuracy.