Data-driven Koopman MPC using mixed stochastic–deterministic tubes

Zhengang Zhong

doi:10.1016/j.jprocont.2025.103533

Data-driven Koopman MPC using mixed stochastic–deterministic tubes

2025-10-21

0

PRE

AI

Z

Zhengang Zhong

E

Ehecatl Antonio del Rio‐Chanona

P

Panagiotis Petsagkourakis *

DOI：10.1016/j.jprocont.2025.103533

Original

Save

Abstract

En 中文

• Introduces mixed stochastic–deterministic tubes to handle model errors and disturbances in data-driven Koopman-based MPC framework. • Uses distributionally robust optimization with Wasserstein ambiguity set and guarantees recursive feasibility and probabilistic closed-loop robustness. • Reduces online computational complexity via offline tube construction.

Data-driven Koopman MPC using mixed stochastic–deterministic tubes

Abstract

Journal

Organization

Cited Papers