Optimization of arbitrary frequency response of microwave bandstop filter

Abstract

The optimization tool developed consists of a combination of evolution strategy with genetic algorithms. The improved algorithm proved it's capability in optimizing a microwave filter application with arbitrary frequency response. The optimization tool is especially tolerant to input parameter definitions and is efficiently capable of overcoming local optima. The algorithm was developed under the constraint of rapid convergence speed which was increased by more than one order of magnitude compared to the evolution-strategy-alone procedure. The developed optimization tool can be applied to a variety of technical problems, where the performance of a technical device has to meet given specifications, and where this performance is defined numerically by the parameters of the device to be optimized.