Constrained Consensus Algorithms With Fixed Step Size for Distributed Convex Optimization Over Multiagent Networks

In this technical note, we are concerned with constrained consensus algorithms for distributed convex optimization with a sum of convex objective functions subject to local bound and equality constraints. In multiagent networks, each agent has its own data on objective function and constraints. All...

Full description

Saved in:
Bibliographic Details
Published inIEEE transactions on automatic control Vol. 62; no. 8; pp. 4259 - 4265
Main Authors Liu, Qingshan, Yang, Shaofu, Hong, Yiguang
Format Journal Article
LanguageEnglish
Published IEEE 01.08.2017
Subjects
Algorithm design and analysis
Constrained consensus
Convergence
Convex functions
Distributed algorithms
distributed optimization
Electronic mail
Linear programming
multiagent network
Optimization
Online AccessGet full text

Cover

More Information
Summary:In this technical note, we are concerned with constrained consensus algorithms for distributed convex optimization with a sum of convex objective functions subject to local bound and equality constraints. In multiagent networks, each agent has its own data on objective function and constraints. All the agents cooperatively find the minimizer, while each agent can only communicate with its neighbors. The consensus of multiagent networks with time-invariant and undirected graphs is proven by the Lyapunov method. Compared with existing consensus algorithms for distributed optimization with diminishing step sizes, the proposed algorithms with fixed step size have better convergence rate. Simulation results on a numerical example are presented to substantiate the performance and characteristics of the proposed algorithms.
ISSN:0018-9286
1558-2523
DOI:10.1109/TAC.2017.2681200