DRUM Community: Institute for Systems Research
http://hdl.handle.net/1903/4375
2015-04-18T08:40:59ZInstances for the Generalized Regenerator Location Problem
http://hdl.handle.net/1903/16328
Title: Instances for the Generalized Regenerator Location Problem
Authors: Chen, Si; Ljubic, Ivana; Raghavan, S.2015-01-01T00:00:00ZA GENERAL FRAMEWORK FOR CONSENSUS NETWORKS
http://hdl.handle.net/1903/16324
Title: A GENERAL FRAMEWORK FOR CONSENSUS NETWORKS
Authors: Somarakis, Christoforos
Abstract: A new framework for the analysis of consensus networks is developed. The theory
consists of necessary and sufficient conditions and it is flexible enough to comprise a variety of
consensus systems. Under mild connectivity assumptions, the discussion ranges from linear, nonlinear, ordinary, functional and leader-follower models. The establishment of explicit estimates on
the rate of convergence is the central objective. Our work extends and unifies past related works in
the literature. Illustrative examples and simulations are presented to outline the theoretical results.2015-03-10T00:00:00ZNonlinear Programming Methods for Distributed Optimization
http://hdl.handle.net/1903/16055
Title: Nonlinear Programming Methods for Distributed Optimization
Authors: Matei, Ion; Baras, John
Abstract: In this paper we investigate how standard nonlinear programming algorithms can be used to solve constrained
optimization problems in a distributed manner. The optimization setup consists of a set of agents interacting through
a communication graph that have as common goal the minimization of a function expressed as a sum of (possibly
non-convex) differentiable functions. Each function in the sum corresponds to an agent and each agent has associated
an equality constraint. By re-casting the distributed optimization problem into an equivalent, augmented centralized
problem, we show that distributed algorithms result naturally from applying standard nonlinear programming tech-
niques. Due to the distributed formulation, the standard assumptions and convergence results no longer hold. We
emphasize what changes are necessary for convergence to still be achieved for three algorithms: two algorithms
based on Lagrangian methods, and an algorithm based the method of multipliers. The changes in the convergence
results are necessary mainly due to the fact that the (local) minimizers of the lifted optimization problem are not
regular, as a results of the distributed formulation. Unlike the standard algorithm based on the method of multipliers,
for the distributed version we cannot show that the theoretical super-linear convergence rate can be achieved.2015-01-01T00:00:00ZTowards a unified theory of consensus
http://hdl.handle.net/1903/15969
Title: Towards a unified theory of consensus
Authors: Somarakis, Christoforos
Abstract: We revisit the classic multi-agent distributed consensus problem under mild connectivity assumptions and non-uniformly bounded weights. The analysis is based on a novel application of the standard results from the non-negative matrix theory. It is a simple, yet unifying, approach that yields generalized results. We apply these results to a wide variety of linear, non-linear consensus and flocking algorithms proposed in the literature and we obtain new conditions for asymptotic consensus. Our framework is developed in both discrete and continuous time. Furthermore we extend the discussion to stochastic settings.2014-10-12T00:00:00Z