Asynchronous Newton-Raphson Consensus for Robust Distributed Convex Optimization
نویسندگان
چکیده
A general trend in the development of distributed 1 convex optimization procedures is to robustify existing algo2 rithms so that they can tolerate the characteristics and condi3 tions of communications among real devices. This manuscript 4 follows this tendency by robustifying a promising distributed 5 convex optimization procedure known as Newton-Raphson 6 consensus. More specifically, we modify this algorithm so 7 that it can cope with asynchronous, broadcast and unreliable 8 communications. We prove the convergence properties of the 9 modified algorithm under the assumption that the local costs are 10 quadratic, and support with numerical simulations the intuition 11 that this robustified algorithm converges to the true optimum as 12 soon as the local costs satisfy some mild smoothness conditions. 13
منابع مشابه
Asynchronous Newton-Raphson Consensus for Distributed Convex Optimization !
We consider the distributed unconstrained minimization of separable convex cost functions, where the global cost is given by the sum of several local and private costs, each associated to a specific agent of a given communication network. We specifically address an asynchronous distributed optimization technique called Newton-Raphson Consensus. Beside having low computational complexity, low co...
متن کاملNewton - Raphson Consensus 1 for Distributed Convex Optimization
5 We address the problem of distributed unconstrained convex optimization under separability assumptions, i.e., the framework 6 where a network of agents, each endowed with a local private multidimensional convex cost and subject to communication constraints, 7 wants to collaborate to compute the minimizer of the sum of the local costs. We propose a design methodology that combines average 8 co...
متن کاملDistributed Newton Methods for Strictly Convex Consensus Optimization Problems in Multi-Agent Networks
Various distributed optimization methods have been developed for consensus optimization problems in multi-agent networks. Most of these methods only use gradient or subgradient information of the objective functions, which suffer from slow convergence rate. Recently, a distributed Newton method whose appeal stems from the use of second-order information and its fast convergence rate has been de...
متن کاملA New Load-Flow Method in Distribution Networks based on an Approximation Voltage-Dependent Load model in Extensive Presence of Distributed Generation Sources
Power-flow (PF) solution is a basic and powerful tool in power system analysis. Distribution networks (DNs), compared to transmission systems, have many fundamental distinctions that cause the conventional PF to be ineffective on these networks. This paper presents a new fast and efficient PF method which provides all different models of Distributed Generations (DGs) and their operational modes...
متن کاملConvergence Analysis of Generalized Iteratively Reweighted Least Squares Algorithms on Convex Function Spaces
The computation of robust regression estimates often relies on minimization of a convex functional on a convex set. In this paper we discuss a general technique for a large class of convex functionals to compute the minimizers iteratively which is closely related to majorization-minimization algorithms. Our approach is based on a quadratic approximation of the functional to be minimized and inc...
متن کامل