Research On 2D Dielectric Target Reconstruction Algorithm Combined With Stochastic Optimization

The classical electromagnetic inversion algorithm establishes the cost function to minimize the difference between the experimental measurement data and the scattering data corresponding to the predicted distribution.The common iterative optimization method is the deterministic optimization algorithm.The deterministic optimization algorithm can effectively converge to the optimal solution under simple and good operating conditions,but in practical application,many factors such as noise interference and the limitation of excitation wave have a bad influence on it.Due to rely too much on gradient calculation,deterministic optimization is easy to fall into local extremum in adverse measurement environment.Based on this,in the framework of microwave imaging,this paper proposes electromagnetic detection of two-dimensional dielectric cylinders based on two kinds of random optimization algorithms,particle swarm optimization(PSO)and bat algorithm(BA),studies the implementation mechanism of the algorithm,and analyzes the performance through numerical simulation experiments.Firstly,according to the basic electromagnetic scattering theory,in the paper we build a two-dimensional transverse magnetic(TM)wave imaging model.The singlefrequency incident wave is incident on the scatterer at multiple angles,and the states of the scatterers in the domain of interest(DOI)are predicted by the echo signal received at the receivers.The electromagnetic integral equation is used to describe the electromagnetic field in the DOI,and the physical parameters of the targets are connected with the electromagnetic field in the observation area.Because the integral formula of electromagnetic scattering is ill-posed and nonlinear,the inverse scattering problem is transformed into an optimization problem and a regularization term is added to make the solution unique.Then,based on the microwave imaging model,the inverse scattering problem of two-dimensional dielectric cylinders is studied by PSO algorithm.The imaging accuracy of dielectric cylinders reconstructed by PSO algorithm under different population size and noise intensity is discussed.Next,the inverse scattering problem of two-dimensional dielectric cylinders based on BA algorithm is proposed.Simulation results verify the accuracy,effectiveness and robustness of PSO and BA algorithm for parameters reconstruction of scattering targets.Finally,this paper proposes to solve the electromagnetic inverse scattering problem based on pixels by using PSO and BA inversion algorithm combined with priori information.In order to improve the computational efficiency of the PSO and BA algorithm,the priori information is integrated into the inversion algorithm.In order to improve the computational efficiency of PSO and BA algorithm,the prior information that the blurred image of the scatterer is obtained by diffraction chromatography(DT)is used to replace a particle in PSO and BA algorithm.The scatterer of image is reconstructed based on pixels to obtain the distribution of the relative permittivity of the scatterer in the DOI.The experimental results show that the PSO and BA algorithm combined with DT is robust,and the hybrid PSO can accurately reconstruct the relative permittivity distribution of multiple isolated scatterers;Hybrid BA algorithm is effective for the contour and position reconstruction of square scatterers and two dielectric cylinders.