Distributed Inexact Computing for Global Constraint-Coupled Optimization in Industrial Cyber-Physical Systems
IEEE Transactions on Industrial Cyber-Physical Systems · 2026 · 4 · 1141-1155
DOI: 10.1109/ticps.2026.3739700 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Industrial Cyber-Physical Systems, 2026. Verfasst von Liang Ran und 4 weiteren Autor:innen. Die Arbeit nennt 46 Literaturangaben. Gefördert von: Fundamental Research Funds for the Central Universities, National Natural Science Foundation of China, Chongqing Science Fund for Distinguished Young Scholars.
Abstract (Originalsprache)
This paper addresses a general distributed constrained optimization problem over networked industrial cyber-physical systems, featuring globally coupled constraints as well as locally edge-based constraints. Each agent in the network privately possesses two general convex objective functions and an additional convex composite term involving a linear mapping. To enable cooperative decision-making towards an optimal solution through local communication, we propose a synchronous distributed inexact multiplier algorithm (DIMA) based on Lagrangian multiplier and proximal splitting techniques, along with its randomized version (RanDIMA) built upon the randomized block-coordinate activation scheme. The proposed algorithms offer several key advantages: (i) they are applicable to a wide range of existing distributed composite optimization problems, including consensus-based unconstrained problems, distributed resource allocation problems, and constrained problems with convex inequality constraints; (ii) each agent is permitted to solve its subproblem inexactly without requiring stringent precision criteria; and (iii) the randomized version does not mandate activation of all agents at every iteration, thereby affording greater flexibility than synchronous algorithms. Theoretically, the convergence of both DIMA and RanDIMA is guaranteed under the same step-size conditions and mild convexity assumptions, without relying on restrictive conditions such as Lipschitz continuity or boundedness of feasible sets. Comparative simulations demonstrate the effectiveness and performance superiority of the proposed methods.
Förderung
- Fundamental Research Funds for the Central Universities
- National Natural Science Foundation of China
- Chongqing Science Fund for Distinguished Young Scholars
Zitieren
APA
Ran, L., Li, H., Zheng, L., Li, J., & Tang, R. (2026). Distributed Inexact Computing for Global Constraint-Coupled Optimization in Industrial Cyber-Physical Systems. IEEE Transactions on Industrial Cyber-Physical Systems, 4, 1141-1155. https://doi.org/10.1109/ticps.2026.3739700
Vancouver
Ran L, Li H, Zheng L, Li J, Tang R. Distributed Inexact Computing for Global Constraint-Coupled Optimization in Industrial Cyber-Physical Systems. IEEE Transactions on Industrial Cyber-Physical Systems. 2026;4:1141-1155. doi:10.1109/ticps.2026.3739700
BibTeX
@article{Ran2026,
author = {Ran, Liang and Li, Huaqing and Zheng, Lifeng and Li, Jun and Tang, Run},
title = {Distributed Inexact Computing for Global Constraint-Coupled Optimization in Industrial Cyber-Physical Systems},
journal = {IEEE Transactions on Industrial Cyber-Physical Systems},
year = {2026},
volume = {4},
pages = {1141--1155},
doi = {10.1109/ticps.2026.3739700}
}
EUVIMED