Skip to content
Preprint

A Koopman--Polynomial Chaos Framework for SMPC of Nonlinear Stochastic Processes

Oct 2026 · 0 citations · 43 references
Engineering Computer Science

TL;DR

A polynomial parametric Koopman operator (PPKO) framework for stochastic model predictive control (SMPC) of nonlinear systems with parametric uncertainty and time-varying stochastic disturbances is developed, showing that the polynomial parameterization introduces no irreducible approximation gap beyond PCE truncation.

Abstract

This paper develops a polynomial parametric Koopman operator (PPKO) framework for stochastic model predictive control (SMPC) of nonlinear systems with parametric uncertainty and time-varying stochastic disturbances. The Koopman operator is parameterized by polynomial chaos expansions (PCEs), enabling data-driven nonlinear dynamics to be integrated with PCE-based uncertainty propagation and control. We show that the polynomial parameterization introduces no irreducible approximation gap beyond PCE truncation. We then derive a condensed deterministic SMPC reformulation that handles both time-invariant parametric uncertainty and time-varying disturbances while retaining an online optimization dimension that depends only on the control horizon and input dimension, independent of the lifted-state and PCE dimensions. The resulting formulation yields convex programs with expectation and second-moment constraints that can be solved using standard optimization methods. Numerical studies on the Duffing oscillator and the electrothermal cooling system demonstrate the accuracy, scalability, and closed-loop performance of the proposed framework.

View source

Similar papers

Open access 2026

Error Bounds for Finite-Dimensional Stochastic Koopman Approximations

Koopman operator theory provides a popular framework for lifting a nonlinear dynamical system to a linear system in a higher dimensional space using a dictionary of observable functions. When a finite-dimensional dictionary is selected, it is important to determine whether the corresponding observable subspace is invar...

S. R. Jafari, H. Modares, F. Yaghmaie · 0 citations
Sep 2026

Physics-Informed Neural Network Surrogates with Polynomial Chaos-Based Uncertainty Propagation for Stochastic Model Predictive Control

An error decomposition is provided for the PINN-PCE surrogate that separates PCE truncation, stochastic quadrature, and PINN approximation errors and it is shown how the surrogate approximation error can be incorporated into tightened probabilistic constraints.

Srimanta Santra, R. Patel, Saikat Mukherjee et al. · 0 citations
Preprint Sep 2026

Contraction-Based Model Predictive Control using Bilinear Koopman Realizations with Proportional Error Bounds

Data-driven model predictive control based on Koopman operator theory is a promising approach for constrained control of nonlinear systems with unknown dynamics. This paper proposes a robust model predictive control framework for such systems using bilinear Koopman realizations with state- and input-dependent proportio...

Yuki Higuchi, Kazuhiro Sato · 0 citations
Preprint Oct 2026

Full-Order Safety for Nonlinear Reduced-Order MPC

Reduced-order models (ROMs) make model predictive control (MPC) of high-dimensional nonlinear systems tractable, but the resulting modeling error can lead to constraint violations on the real system. We introduce CoRE-MPC, a robust MPC formulation that plans with a nonlinear ROM and guarantees recursive feasibility and...

Luis A. Pabon, Johannes Köhler, Revant Kasichainula et al. · 0 citations
Preprint Oct 2026

Koopman--Lyapunov and Koopman--Riccati Controller Synthesis

While the Koopman operator-based lifted representations of nonlinear autonomous and controlled systems have been well covered in the literature, the direct controller synthesis problem largely remains open. As the controlled system is generally a bilinear (instead of fully linear) Koopman model, controller synthesis ap...

Wen-Tao Tang · 0 citations
Open access Sep 2026

DGA: a differential relaxation of genetic algorithms for receding-horizon control of nonlinear systems

The growing complexity of industrial cyber-physical systems demands control strategies capable of handling strong nonlinearities, tight actuator constraints, and competing objectives under real-time requirements. Although Genetic Algorithms (GAs) are widely used in this class of problems, their non-differentiable opera...

Ali Deeb, V. Khokhlovskiy, Viacheslav Shkodyrev · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.