Efficient GA-inspired macro-modeling of general LTI multi-port systems

Abstract

A numerically robust sampling and rational fitting method is introduced, that models the entire state-space matrix of multiple-input-multiple-output (MIMO) linear time-invariant (LTI) systems. The algorithm adaptively builds an accurate rational pole-residue model, based on a minimal set of support samples. During the modeling process, no prior knowledge of the system's dynamics is required. The "survival-of-the-fittest" principle of a genetic algorithm (GA) provides a reliable way to detect convergence of the modeling process.