A Parallel-in-Time Newton's Method for Nonlinear Model Predictive Control

Model predictive control (MPC) is a powerful framework for optimal control of dynamical systems. However, MPC solvers suffer from a high computational burden that restricts their application to systems with low sampling frequencies. This issue is further amplified in nonlinear and constrained system...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE transactions on control systems technology S. 1 - 8
Hauptverfasser: Iacob, Casian, Abdulsamad, Hany, Sarkka, Simo
Format: Journal Article
Sprache:Englisch
Veröffentlicht: IEEE 2025
Schlagworte:
Computational efficiency
Constrained nonlinear optimization
Convergence
Convex functions
Dynamic programming
Heuristic algorithms
IP networks
model predictive control (MPC)
Newton method
Optimal control
Optimization
parallel computation
Planning
ISSN:1063-6536, 1558-0865
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Model predictive control (MPC) is a powerful framework for optimal control of dynamical systems. However, MPC solvers suffer from a high computational burden that restricts their application to systems with low sampling frequencies. This issue is further amplified in nonlinear and constrained systems that require nesting MPC solvers within iterative procedures. In this brief, we address these issues by developing parallel-in-time algorithms for constrained nonlinear optimization problems that take advantage of massively parallel hardware to achieve logarithmic computational time scaling over the planning horizon. We develop time-parallel second-order solvers based on interior point (IP) methods and the alternating direction method of multipliers (ADMM), leveraging fast convergence and lower computational cost per iteration. The parallelization is based on a reformulation of the subproblems in terms of associative operations that can be parallelized using the associative scan algorithm. We validate our approach on numerical examples of nonlinear and constrained dynamical systems.
ISSN:1063-6536
1558-0865
DOI:10.1109/TCST.2025.3589409