Genetic algorithms applied to microwave filters optimization and design

This paper presents improvements in genetic algorithms for the optimization of microwave circuits. Several fitness functions are studied and implemented in the optimization process. They allow to quantify the importance of the specifications and of the solution approach with respect to the ideal circuit performance for a bandpass filter. The performance of the optimization method with different fitness functions is compared for the design of a Ku-band two-pole rectangular iris coupled waveguide bandpass filter. Copyright © (2008) by the Electromagnetics Academy.