Experimental demonstration of multi-agent learning and planning under uncertainty for persistent missions with automated battery management

N. Kemal Ure, Tuna Toksoz, Girish Chowdhary, Joshua Redding, Jonathan P. How, Matthew A. Vavrina, John Vian

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper presents algorithms and ight test results for multi-agent cooperative planning problems in presence of state-correlated uncertainty.An online learning and planning framework is used to address the problem of improving planner performance for missions with state-dependent uncertain agent health dynamics. The framework includes a previously introduced Decentralized Multi-agent Markov decision process (Dec-MMDP) as an online planning algorithm that is scalable in number of agents, and Incremental Feature Discovery (iFDD) which is a compact and fast learning algorithm for estimating parameters of a state-correlated uncertainty model. In combination, this architecture yield an integrated learning-planning algorithm where the planning performance improves as uncertainty is reduced through learning. The presented algorithms are validated in a persistent search and track scenario with a novel automated battery swapping/recharging system that enables the UAVs to collaboratively track targets over durations that are significantly larger than individual vehicle endurance with a single battery. The results indicate that the architecture can be used as an computationally effcient solution to multi-agent uncertain cooperative planning problems.

Original languageEnglish (US)
Title of host publicationAIAA Guidance, Navigation, and Control Conference 2012
PublisherAmerican Institute of Aeronautics and Astronautics Inc.
ISBN (Print)9781600869389
DOIs
StatePublished - 2012
EventAIAA Guidance, Navigation, and Control Conference 2012 - Minneapolis, MN, United States
Duration: Aug 13 2012Aug 16 2012

Publication series

NameAIAA Guidance, Navigation, and Control Conference 2012

Conference

ConferenceAIAA Guidance, Navigation, and Control Conference 2012
CountryUnited States
CityMinneapolis, MN
Period8/13/128/16/12

ASJC Scopus subject areas

  • Aerospace Engineering
  • Control and Systems Engineering
  • Electrical and Electronic Engineering

Fingerprint Dive into the research topics of 'Experimental demonstration of multi-agent learning and planning under uncertainty for persistent missions with automated battery management'. Together they form a unique fingerprint.

Cite this