Abstract | ||
---|---|---|
We consider the problem of steering the state of a group of agents with integrator dynamics to the solution of an optimization problem in the form of a sum of strictly convex separable functions, each available to an individual agent, with shared linear constraints. Under the assumption that the agents are capable of communicating over a network prescribed by a directed graph, not necessarily weight-balanced, we provide a procedure for designing a class of gradient-free distributed extremum seeking control laws that achieve this objective. The novelty of this procedure is in taking advantage of Lie brackets to generate a distributed dynamics that approximates the trajectories of the corresponding saddle-point dynamics, which is not necessarily distributed. Several examples illustrate our results. |
Year | Venue | Field |
---|---|---|
2017 | 2017 IEEE 56TH ANNUAL CONFERENCE ON DECISION AND CONTROL (CDC) | Saddle,Mathematical optimization,Computer science,Separable space,Integrator,Directed graph,Convex function,Linear programming,Optimization problem,Trajectory |
DocType | ISSN | Citations |
Conference | 0743-1546 | 1 |
PageRank | References | Authors |
0.37 | 0 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Simon Michalowsky | 1 | 6 | 3.21 |
Bahman Gharesifard | 2 | 340 | 26.54 |
Christian Ebenbauer | 3 | 200 | 30.31 |