This paper studies the application of the blended dynamics approach towards the distributed optimization problem . The benefits include (i) individual cost functionneed not be convex as long as the global cost function is strongly convex and (ii) the convergence rate of the distributed algorithm is arbitrarily close to the convergence rates of the centralized one . Two particular continuous-time algorithms are presented using the proportional-integral-type couplings. Onehas benefit of `initialization-free,’ so that agents can join or leave thenetwork during the operation. The other one has the minimal amount of communication information. One of the algorithms has the minimum amount of information .

Author(s) : Seungjoon Lee, Hyungbo Shim

Links : PDF - Abstract

Code :

https://github.com/nhynes/abc


Coursera

Keywords : convergence - distributed - rate - algorithms - cost -

Leave a Reply

Your email address will not be published. Required fields are marked *