Data-driven robust reference governor for constrained nonlinear systems

Yao Shi

doi:10.1080/00207179.2025.2575289

Data-driven robust reference governor for constrained nonlinear systems

2025-10-01

0

PRE

AI

DOI：10.1080/00207179.2025.2575289

Original

Original request for help

Save

Abstract

En 中文

We present a data-driven robust reference governor (DRRG) framework that handles output constraints in nonlinear systems with unknown dynamics and uncertainties. By incorporating the Koopman operator, the DRRG first maps nonlinear system dynamics into a higher-dimensional linear space, allowing for more effective data-driven prediction. Subsequently, we develop a data-driven maximal admissible set (MAS) and a min-max optimisation problem using the high-dimensional data-driven predictors to compute optimal reference signals that are resilient to all possible realizations of uncertainties. The proposed DRRG method provides a practical solution to improving robustness against uncertainties while retaining the existing control architectures. Numerical simulations are presented to demonstrate the effectiveness of the approach.

Keywords:

Reference governor

data-driven control

nonlinear system

maximal admissible set

min-max optimisation

Data-driven robust reference governor for constrained nonlinear systems

Abstract

Journal

Organization