A differential evolution algorithm for the extraction of complex natural resonance frequencies of electromagnetic targets

Abstract

This paper presents a differential evolution algorithm in order to find unique resonance frequencies of an electromagnetic target in the resonance scattering region. These frequencies are estimated from the roots of Laplace transform of a specially designed incident signal. The parameters of the signal are computed with an intelligent differential evolution (DE) algorithm. The algorithm searches for minimization of the scattered signal's energy in late-time region, which is main fitness function in the algorithm. The proposed algorithm is demonstrated for a scattered signal of a dielectric sphere having several poles. The acquired pole results show very good agreement with theoretical expectations. Besides, the differential evolution algorithm has higher accuracy as compared to a similar method, which utilizes from genetic algorithm. © 2011 Springer-Verlag.