矩量法中引入压缩感知求解三维电磁散射问题

Application of compressed sensing to solution of three-dimensional electromagnetic scattering problems by MoM

在解决三维电磁散射问题时,随着入射波角度的变化,传统的矩量法需针对每个角度分别计算该入射波照射下散射体表面的电流系数。对于求解宽角度下的电磁散射问题,该算法需反复运用迭代求解,运算量大,效率相对较低。为此,引入压缩感知技术来求解宽角度下的三维电磁散射问题,利用稀疏转换基和观测矩阵对数据进行稀疏表示和观测,通过数次观测所得的值再采用恢复算法即可重构出所需全部入射角度下的电流系数值。利用该方法计算所得结果与传统算法得到的结果一致,在不影响计算精度的同时可减少运算时间,降低计算复杂度。

With a change in the incident angle, the traditional method of moments （MoM） needs to calcu- late current coefficients on scatterer＇s surface each time in solving three-dimensional electromagnetic scattering problems. When solving electromagnetic scattering problems over wide angles, this method takes iterative oper ations over every angle, the amount of computation is great and the efficiency is low. So the compressed sensing （CS） is introduced into solution of three-dimensionai electromagnetic scattering problems over wide angles. Data can be sparely expressed and measured by a sparse transformation basis and a measurement matrix, then each group of current coefficients needed over any incident angle could be reconstructed through these several meas- urements by a recovery algorithm. In contrast with the traditional MoM, this new method can greatly reduce the computational time and complexity with computation accuracy not to be affected.