An approximate dual subgradient algorithm for distributed non-convex constrained optimization

Minghui Zhu and Sonia Martínez
IEEE Transactions on Automatic Control, 58 (6) (2013) 1534-1539

Abstract:

We consider a multi-agent optimization problem where agents aim to cooperatively minimize a sum of local objective functions subject to a global inequality constraint and a global state constraint set. In contrast to existing papers, we do not require that the objective, constraint functions, and state constraint sets are convex. We propose a distributed approximate dual subgradient algorithm to enable agents to asymptotically converge to a pair of approximate primal-dual solutions over dynamically changing network topologies. Convergence can be guaranteed provided that the Slater's condition and strong duality property are satisfied.

File: main.pdf (Extended version, also available at Arxiv)

Bib-tex entry:

@article{MZ-SM:13,
author = {M. Zhu and S. Mart{\'\i}nez},
title = {An approximate dual subgradient algorithm for distributed non-convex constrained optimization},
journal= {IEEE Transactions on Automatic Control},
volume = 58,
number = 6,
pages = {1534--1539},
year = 2013,
note = {Available at http://arxiv.org/abs/1010.2732}
}