Model order reduction for discrete time-delay systems with inhomogeneous initial conditions

TL;DR

This paper introduces two novel model order reduction techniques for discrete time-delay systems with inhomogeneous initial conditions, utilizing Walsh functions and Laguerre polynomials to improve efficiency and accuracy.

Contribution

The paper develops new reduction methods that account for inhomogeneous initial conditions using Walsh functions and a decomposition approach with Gramians, enhancing existing techniques.

Findings

Reduced models preserve key Walsh coefficients.

The methods effectively handle inhomogeneous initial conditions.

Numerical examples demonstrate improved efficiency and accuracy.

Abstract

We propose two kinds of model order reduction methods for discrete time-delay systems with inhomogeneous initial conditions. The peculiar properties of discrete Walsh functions are directly utilized to compute the Walsh coefficients of the systems, and the projection matrix is defined properly to generate reduced models by taking into account the non-zero initial conditions. It is shown that reduced models can preserve some Walsh coefficients of the expansion of the original systems. Further, the superposition principle is exploited to achieve a decomposition of the original systems, and a new definition of Gramians is proposed by combining the individual Gramians of each subsystem. As a result, the balanced truncation method is applied to systems with inhomogeneous initial conditions. We also provide a low-rank approximation to Gramians based on the discrete Laguerre polynomials, which enables an efficient execution of our approach. Numerical examples confirm the feasibility and effectiveness of the proposed methods.