基于SOM算法的等离子体参数反演方法

      A robust inversion method of plasma parameters based on SOM algorithm

      • 摘要: 电磁逆散射在无损检测和生物医学成像中有着广泛的应用。本文将常规介质的逆散射成像问题扩展至色散介质,研究等离子体散射场与等离子体参数之间的非线性关系,提出了一种基于子空间优化法(subspace-based optimization method, SOM)和数据后处理的逆散射方法,用于重构由德鲁(Drude)模型描述的色散介质等离子体。针对碰撞频率与等离子体介电常数之间非线性关系导致的多解问题,对碰撞频率重构公式进行了修正以确定唯一解。此外,为了改善碰撞频率的反演精度,利用精度较高的等离子体电子密度重构结果改进反演方法。仿真结果验证了上述方法的快速收敛性和对噪声的鲁棒性。本文研究可以为等离子体参数重构提供一种高效且稳定的方法。

         

        Abstract: Electromagnetic inverse scattering(EMIS) has been widely used in nondestructive testing and biomedical imaging. In this paper, the inverse scattering imaging problem of conventional media is extended to the inverse scattering problem of dispersive media, and the nonlinear relationship between plasma scattering field and plasma parameters is studied. An inverse scattering method based on subspace-based optimization method (SOM) and postprocessing of data is proposed to reconstruct the dispersive medium plasma described by Drude model. For the multi-solution problem caused by the nonlinear relationship between the collision frequency and the plasma dielectric constant, the collision frequency reconstruction formula is modified to determine the unique solution. In addition, in order to improve the inversion accuracy of collision frequency, the inversion method is improved by using the reconstruction results of plasma electron density with high accuracy. Simulation results verify the fast convergence and robustness to noise of the proposed method. This study can provide an efficient and stable method for plasma parameter reconstruction.

         

      /

      返回文章
      返回