A Framework for Distributed Personalized Optimization in Cyber-Physical Systems via Multi-Parameter Fractional Dynamics

This paper aims at solving the distributed personalized optimization problem in cyber-physical systems by combining the fractional dynamics. To fully utilize the advantages of the basic primal-dual method and its different variants, several parameters are introduced, which brings a unified framework...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE transactions on industrial cyber-physical systems Jg. 3; S. 549 - 558
Hauptverfasser: Ni, Xintong, Wei, Yiheng, Tao, Meng, Hua, Liang, Cao, Jinde
Format: Journal Article
Sprache:Englisch
Veröffentlicht: IEEE 2025
Schlagworte:
Convergence
Cyber-physical systems
Distributed personalized optimization
Eigenvalues and eigenfunctions
Federated learning
Fractional calculus
Heuristic algorithms
Lagrangian functions
Linear programming
multi-parameter system
Optimization
Topology
unified framework
ISSN:2832-7004, 2832-7004
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:This paper aims at solving the distributed personalized optimization problem in cyber-physical systems by combining the fractional dynamics. To fully utilize the advantages of the basic primal-dual method and its different variants, several parameters are introduced, which brings a unified framework. Along with the continuous time algorithms, the discrete time algorithms are constructed. The convergence is analyzed by using the Lyapunov stability theory. To demonstrate the effectiveness and efficiency of the elaborated algorithms, a series of examples are provided.
ISSN:2832-7004
2832-7004
DOI:10.1109/TICPS.2025.3615894