Comparison of particle swarm optimization and self-adaptive dynamic differential evolution for the imaging of a periodic conductor

Abstract

The application of two techniques to reconstruct the shape of a two-dimensional periodic perfect conductor from mimic the measurement data is presented. A periodic conducting cylinder of unknown periodic length and shape scatters the incident wave in half-space and the scattered field is recorded outside. After an integral formulation, the microwave imaging is recast as a nonlinear optimization problem; a cost functional is defined by the norm of a difference between the measured scattered electric fields and the calculated scattered fields for an estimated shape of a conductor. Thus, the shape of conductor can be obtained by minimizing the cost function. In order to solve this inverse scattering problem, transverse magnetic (TM) waves are incident upon the objects and two techniques are employed to solve these problems. The first is based on an particle swarm optimization (PSO) and the second is a self-adaptive dynamic differential evolution (SADDE). Both techniques have been tested in the case of simulated mimic the measurement data contaminated by additive white Gaussian noise. Numerical results indicate that the SADDE algorithm is better than the PSO in reconstructed accuracy and convergence speed.