Data-driven balanced truncation for second-order systems via the approximate Gramians

TL;DR

This paper introduces a data-driven balanced truncation method for second-order systems that uses frequency domain measurements and approximate Gramians, enabling nonintrusive model reduction with improved efficiency.

Contribution

It develops a structure-preserving, nonintrusive balanced truncation approach for second-order systems based on sample data and approximate Gramians, including low-rank solutions for large datasets.

Findings

Effective model reduction demonstrated on numerical examples.

Speed-up achieved via low-rank Sylvester equation solutions.

Method preserves second-order system structure.

Abstract

This paper studies the data-driven balanced truncation (BT) method for second-order systems based on the measurements in the frequency domain. The basic idea is to approximate Gramians used the numerical quadrature rules, and establish the relationship between the main quantities in the procedure of BT with the sample data, which paves the way for the execution of BT in a nonintrusive manner. We construct the structure-preserving reduced models approximately based on the samples of second-order systems with proportional damping, and provide the detailed execution of the data-driven counterpart of BT in real-value arithmetic. The low-rank approximation to the solution of Sylvester equations is also introduced to speed up the process of the proposed approach when a large amount of samples involved in the modeling. The performance of our approach is illustrated in detail via two numerical examples.