On the Analysis of Suboptimal Nonlinear Model Predictive Control Schemes Without Terminal Constraints

MCML Authors

Christian Fiedler

Dr.

→ Group Massimo Fornasier
 Applied Numerical Analysis

Abstract

Model predictive control (MPC) – one of the most important control methodologies, with many real-world applications and being the focus of intense research efforts – is based on repeatedly solving open-loop optimal control problems. Since this requires in general online numerical optimisation, in practice the optimal control problems are solved only up to some suboptimality, though many analyses do not take this into account. Motivated by this gap, we prove performance bounds and stability results under a given level of suboptimality for nonlinear model predictive control schemes without terminal constraints or costs.

article

---

IEEE Control Systems Letters

9. Sep. 2025.

---

Authors

C. Fiedler • S. Trimpe

---

Links

DOI

---

Research Area

 A2 | Mathematical Foundations