Application of the Cauchy method to genetic algorithms for broadband antenna design

The concept of model-based parameter estimation (MBPE) has been extensively used in computational electromagnetics as a means to reduce the computation cost of generating broadband data. The rational function is commonly used to model the input impedance of an antenna. One robust method to derive the coefficients of the rational function model is the Cauchy method. Genetic algorithms (GA) have been extensively used for the design optimization of electromagnetic structures. The application of MBPE to the GA optimization of broadband system appears very natural. We apply the Cauchy method to GA in order to speed up the computation time in designing broadband antennas. The Cauchy method can reduce the time for simulating the input impedance in the GA process. A differential RMS error is proposed as the stopping criterion for arriving at a sufficiently accurate Z_in. Using the developed algorithm, two broadband antennas are designed. A sampling gain metric is proposed and found to be 12.4 and 13.6 in the two test cases.