DMCA
A Model of Adaptation in Collaborative Multi-Agent Systems (2004)
| Venue: | ADAPTIVE BEHAVIOR |
| Citations: | 7 - 0 self |
BibTeX
@INPROCEEDINGS{Lerman04amodel,
author = {Kristina Lerman},
title = {A Model of Adaptation in Collaborative Multi-Agent Systems},
booktitle = {ADAPTIVE BEHAVIOR },
year = {2004},
pages = {187--197},
publisher = {}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
Adaptation is an essential requirement for autonomous agent systems functioning in uncertain dynamic environments. Adaptation allows agents to change the bez vior in orde to improve the oveG88 syste peKJJ9k3K7G We deG8G7 e a geMNKG me hanism for adaptation in multi-age tsysteN in whichage ts modify theH be havior in reM onse to change inthe e nvironme t or actions ofothe age ts.The age tse stimate the globalstate of the syste from local obsez ations and adjustthes actions accordingly. We deNK e a matheNk3zJ9 mode that detkKH e s the colle9k3 e beH vior of such adaptive systeMJ The modez consisting ofcouple Rate Equations, gove rns howthe colleJJk e beM vior change intime We apply the mo de to study collaboration in a group ofmobile robots. The syste we study is an adaptive version of the collaborative stickpulling in a group of robotse xamine in deM88 ine arlie works (Ijsp espk 2001;Le1;kG 2001). In adaptive stickpulling, robotsekMKzK9 the numbe r of robots and sticks inthe syste and adjustthes individual beM vior so as to improve collezk3 e pe rformance We solve the matheMk3M8G mode and show that adaptation improve s colle7Mk e performance for all parameter values.
Keyphrases
collaborative multi-agent system group ofmobile robot rem onse adaptive version ofcouple rate equation adaptive stickpulling adjustthes action collaborative stickpulling multi-age tsysten mathemk3m8g mode robotse xamine age tse parameter value essential requirement bez vior autonomous agent system local obsez ations mathenk3zj9 mode colle7mk performance collezk3 pe rformance whichage t age t oveg88 syste inthe nvironme adaptive systemj dem88 ine arlie uncertain dynamic environment
