Image reconstruction for a partially immersed imperfectly conducting cylinder by genetic algorithm

Abstract

This article presents a computational approach to the imaging of a partially immersed imperfectly conducting cylinder. An imperfectly conducting cylinder of unknown shape and conductivity scatters the incident transverse magnetic (TM) wave in free space while the scattered field is recorded outside. Based on the boundary condition and the measured scattered field, a set of nonlinear integral equations, and the inverse scattering problem are reformulated into an optimization problem. We use genetic algorithm (GA) to reconstruct the shape and the conductivity of a partially immersed imperfectly conducting cylinder. The genetic algorithm is then used to find out the global extreme solution of the cost function. Numerical results demonstrated that, even when the initial guess is far away from the exact one, good reconstruction can be obtained. In such a case, the gradient-based methods often get trapped in a local extreme. In addition, the effect of random noise on the reconstruction is investigated.

Electromagnetic imaging of buried perfectly conducting cylinder targets using the dynamic differential evolution

“…The disturbance vector V due to the mutation mechanism consists of parameter vector X g j , the best particle X g best , and two randomly selected vectors. When compared with typical DE [22], the convergence can be accelerated by (8). iv.…”

“…One way for solving the nonlinear inverse problem is solving the forward scattering problem iteratively to minimize an error function known as the cost function. This function represents the error between the measured scattered fields and the simulated fields during the updates of the evolving objects in each inversion [2][3][4][5][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21]. Recently, several papers for inverse scattering problems have been published on the subject of 2D object about deal with shape reconstruction problems by using Gauss-Newton method [5], genetic algorithms (GAs) [7][8][9][10][11], differential evolution (DE) [12][13][14][15][16], particle swarm optimization (PSO) [16][17][18][19], Neural network [20], and level-set algorithm [21].…”

Computational Approach Based on a Differential Evolution with Self-Adaptive Concept for Microwave Imaging of Two-Dimensional Inverse Scattering Problem

“…These algorithms, which are based on stochastic strategies, offer advantages relative to local inversion algorithms, including strong search ability simplicity, robustness, and insensitivity to 452 C.-C. Chiu and C.-H. Sun ill-posedness. In contrast to traditional computation systems, evolutionary computation (Sun et al, 2010a(Sun et al, , 2010c(Sun et al, , 2008Rekanos, 2008;Semnani et al, 2009Semnani et al, , 2010Chien et al, 2009;Donelli et al, 2006;Chiu et al, 2009;Chien & Chiu, 2005) provides a more robust and efficient approach for solving inverse scattering problems. Particle swarm optimization (PSO) has proven to be a useful method of optimization for difficult and discontinuous multidimensional engineering problems (Poli et al, 2010).…”

Time Domain Inverse Scattering of Buried Inhomogeneous Dielectric Cylinders

“…In general, they tend to get trapped in local minima when the initial trial solution is far from the exact one. Thus, some populationbased stochastic methods, such as genetic algorithm (GA) [6][7][8][9]18], differential evolution (DE) [11,[19][20][21] particle swarm optimization (PSO) [12,13,[22][23][24][25], are proposed to search the global extreme of the inverse problems to overcome the drawback of the deterministic methods.…”

“…Most of the inversion techniques are investigated for microwave imaging by time-harmonic electromagnetic (EM) wave illumination [3][4][5][6][7][8][9][10][11][12][13]. However, the application of wideband incidents is important for microwave imaging because the available information content about unknown objects is more than single frequency scattering data alone.…”