Combining the Genetic Algorithm Approach and the Model-Based Parameter Estimation into an Adaptive Frequency Sampling Algorithm

In this article, we present a new algorithm for the adaptive frequency sampling of passive microwave devices and networks. The algorithm combines the survival-of-the-fittest principle of the genetic algorithm with the model-based parameter estimation. The sampling algorithm does not need any information about derivatives or continuity of the sampled function. Presented application examples show good efficiency and robustness of the algorithm. The adaptive sampling algorithm leads to a significant reduction of simulation time since it chooses only a minimum of sampling points to model the response of the device under test.