Intelligent Distributed Multi-Agent Systems
Abstract
This project aims to develop new tools for analyzing distributed dynamical multiagent systems and for controlling them. We propose to develop distributed algorithms with only limited information transferred between neighboring agents, the motivation being to significantly reduce communication cost and improve resilience and possibly security. We will exploit recent advances in reaching a consensus with limited information which enable each agent in a vector-valued consensus process to transmit only reduced-dimension information, with original states being unobservable. We propose to address the question of how to reliably perform computations in a distributed manner to a large class of consensus-based problems including convex-nonconvex optimization, determining shared fixed points of families of nonlinear functions, solving a system of linear equations, and estimation of linear-nonlinear systems. Built upon the reduced-dimension consensus, we will appeal to the ideas of network redundancy and objective redundancy. We propose to develop efficient, reliable, and optimal control for a dynamical system using distributed feedback, which is arguably the most fundamental problem in distributed feedback control. The type of distributed algorithms this project aims to develop will directly impact any real-world application area where distributed control, estimation and optimization algorithms are needed.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Feb 29, 2024
- Source ID
- FA95502310175
Entities
People
- Ji Liu
Organizations
- Air Force Office of Scientific Research
- Research Foundation for the State University of New York
- United States Air Force