Analysis and Design of Large-Scale Multi-Agent Systems via Continuum Spatial Approximations

reportActive / Technical Report | Accesssion Number: AD1144435 | Open PDF

Abstract:

This research effort has been led by Dr. Martinez at UC San Diego. The work has had particular impact on the development of a science of autonomy, by focusing on the analysis and design of algorithms for the coordination of very large-scale (spatial) networks. The development of efficient algorithms for large-scale systems remains to be challenging and makes the operation of large networked systems often intractable. Two of the main approaches considered in the literature to deal with large-scale systems have been: a) the direct adoption of algorithms originally developed for small scale multi-agent systems to large ensembles, or b) the discretization of algorithms developed for large scale systems for implementation of small-scale groups. Neither of these translations is optimal, as by missing the relationship between the two we may obtain algorithms that are i) unnecessarily conservative, ii) do not capture or exploit large-scale effects (single agents are negligible) and their impact on performance, and iii) do not capture the under-lying geometric properties (or structure) of interest that characterize large dynamic systems.

Security Markings

DOCUMENT & CONTEXTUAL SUMMARY

Distribution Code:
A - Approved For Public Release
Distribution Statement: Public Release

RECORD

Collection: TRECMS
Identifying Numbers
Subject Terms