An improved evolution strategy and its application to inverse scattering in microwave imaging

This paper presents an efficient and reliable inversion methodology for microwave imaging based on an improved evolution strategy. The proposed non-parametric mutation process, adaptive tradeoff between convergence speed and solution quality during iterations, flexible hybridization strategy, and non-revisiting property are demonstrated to have excellent performance for solving inverse problems with expensive computational evaluations. The numerical results and comparisons with other well developed algorithms as reported positively confirm the feasibilities and advantages of the proposed algorithm.