Informativity for data-driven model reduction through interpolation

TL;DR

This paper introduces a data-driven interpolatory model reduction method that computes transfer function values directly from time-domain data without system identification, using a data informativity framework to ensure accurate reduced models.

Contribution

It presents a novel data informativity approach that characterizes when all systems fitting the data share the same transfer function value, enabling classical interpolation for model reduction.

Findings

Method computes transfer function values from data without explicit system identification.

Framework guarantees conditions under which reduced models match original system behavior.

Electrical circuit example demonstrates practical applicability.

Abstract

A method for data-driven interpolatory model reduction is presented in this extended abstract. This framework enables the computation of the transfer function values at given interpolation points based on time-domain input-output data only, without explicitly identifying the high-order system. Instead, by characterizing the set of all systems explaining the data, necessary and sufficient conditions are given under which all systems in this set share the same transfer function value at a given interpolation point. After following this so-called data informativity perspective, reduced-order models can be obtained by classical interpolation techniques. An example of an electrical circuit illustrates this framework.