On distributed optimization under inequality constraints via Lagrangian primal-dual methods

We consider a multi-agent convex optimization problem where agents are to minimize a sum of local objective functions subject to a global inequality constraint and a global constraint set. To deal with this, we devise a distributed primal-dual subgradient algorithm which is based on the characteriza...

Full description

Saved in:
Bibliographic Details
Published inProceedings of the 2010 American Control Conference pp. 4863 - 4868
Main Authors Minghui Zhu, Martínez, Sonia
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.06.2010
Subjects
Algorithm design and analysis
Communication system control
Constraint optimization
Distributed algorithms
Distributed computing
Lagrangian functions
Multiagent systems
Network topology
Optimization methods
Utility programs
Online AccessGet full text
ISBN9781424474264
1424474264
ISSN0743-1619
DOI10.1109/ACC.2010.5530903

Cover

More Information
Summary:We consider a multi-agent convex optimization problem where agents are to minimize a sum of local objective functions subject to a global inequality constraint and a global constraint set. To deal with this, we devise a distributed primal-dual subgradient algorithm which is based on the characterization of the primal-dual optimal solutions as the saddle points of the Lagrangian function. This algorithm allows the agents to exchange information over networks with time-varying topologies and asymptotically agree on a pair of primal-dual optimal solutions and the optimal value.
ISBN:9781424474264
1424474264
ISSN:0743-1619
DOI:10.1109/ACC.2010.5530903