摘要
An electromagnetic （EA1） scattering model for layered media covered by a 3D infinite rough surface and the corresponding inversion technique are investigated. The work aims at remote sensing the surface roughness and dielectric constant for different depths of bear soil through radar measurement data. The forward problem is carried out by the wave decomposition method. The small perturbation method （SPM） and EM boundary conditions are employed to solve the integral equations introduced by the wave decomposition method. The second-order SPM solution of the scatteringfield is involved in the computation of the forward problem for the first time. The backseattering coefficients of multiple frequencies, multiple angles and multiple polarizations are employed to create a nonlinear optimization problem. A genetic algorithm is introduced to help the inversion procedure approaeh to the global minimum of the cost function. Examples are carried out to validate the inversion technique. The inverion results show good agreement with the forward problem with given parameters and pose good tolerance to the input data with the additive white Gaussian noise.
An electromagnetic （EA1） scattering model for layered media covered by a 3D infinite rough surface and the corresponding inversion technique are investigated. The work aims at remote sensing the surface roughness and dielectric constant for different depths of bear soil through radar measurement data. The forward problem is carried out by the wave decomposition method. The small perturbation method （SPM） and EM boundary conditions are employed to solve the integral equations introduced by the wave decomposition method. The second-order SPM solution of the scatteringfield is involved in the computation of the forward problem for the first time. The backseattering coefficients of multiple frequencies, multiple angles and multiple polarizations are employed to create a nonlinear optimization problem. A genetic algorithm is introduced to help the inversion procedure approaeh to the global minimum of the cost function. Examples are carried out to validate the inversion technique. The inverion results show good agreement with the forward problem with given parameters and pose good tolerance to the input data with the additive white Gaussian noise.
基金
Supported by the National High-Technology Research and Development Program of China under Grant No 2009AA12Z132, and the National Nature Science Foundation of China under Grant No 60890071.
