Microwave imaging of a perfectly conducting cylinder using a real-coded genetic algorithm

The paper presents a novel approach for microwave imaging of a perfectly conducting cylinder in free space using measured scattered data, with or without the effect of random noise. The contour of the cylinder is denoted by a shape function which is approximated by trigonometric series. The scattering problem is solved using point-matching method. The inverse problem is cast into a restrained optimisation problem. The coefficients of the trigonometric series are the variables to be optimised. The relative error between the measured scattered electric field and the simulated one is the object function to be minimised. The optimisation problem is then solved by a real-coded genetic algorithm (RGA). Some numerical examples are presented to examine the performance of this method. The performance of this algorithm is also compared with that of the Newton-Kantorivitch method.