An improved algorithm for microwave imaging of parallel perfectly conducting cylinders

The electromagnetic inverse scattering problem has been one of the most challenging research topics due to its considerable practical importance in various areas of technology. A variety of algorithms have been previously proposed. In this paper a novel algorithm, the real-coded genetic algorithm-Newton-Kantorivitch method (RGA-NKM), which aims to improve the converging performance of the RGA with the help of NKM for microwave imaging of parallel perfectly conducting cylinders, is proposed and presented. The main idea of the algorithm is to perform a Newton-Kantorivitch type search for the local optimum after the genetic operations in each genetic evolution to improve the local search ability of RGA. Numerical results and comparisons with both RGA and NKM demonstrate that although the simplicity of RGA is lost, the convergence is sped up significantly while the other merits of RGA are retained.